{"id":7093,"date":"2026-10-13T09:05:56","date_gmt":"2026-10-13T07:05:56","guid":{"rendered":"https:\/\/www.cim-atlantique.com\/deeper\/"},"modified":"2026-10-13T09:28:20","modified_gmt":"2026-10-13T07:28:20","slug":"deeper","status":"publish","type":"page","link":"https:\/\/www.cim-atlantique.com\/fr\/deeper\/","title":{"rendered":"Deeper"},"content":{"rendered":"<div class=\"fusion-fullwidth fullwidth-box fusion-builder-row-1 fusion-flex-container hundred-percent-fullwidth non-hundred-percent-height-scrolling\" style=\"--awb-border-radius-top-left:0px;--awb-border-radius-top-right:0px;--awb-border-radius-bottom-right:0px;--awb-border-radius-bottom-left:0px;--awb-padding-top:0px;--awb-padding-right:0px;--awb-padding-bottom:0px;--awb-padding-left:0px;--awb-flex-wrap:wrap;\" ><div class=\"fusion-builder-row fusion-row fusion-flex-align-items-flex-start fusion-flex-content-wrap\" style=\"width:104% !important;max-width:104% !important;margin-left: calc(-4% \/ 2 );margin-right: calc(-4% \/ 2 );\"><div class=\"fusion-layout-column fusion_builder_column fusion-builder-column-0 fusion_builder_column_1_1 1_1 fusion-flex-column\" style=\"--awb-bg-size:cover;--awb-width-large:100%;--awb-margin-top-large:0px;--awb-spacing-right-large:1.92%;--awb-margin-bottom-large:20px;--awb-spacing-left-large:1.92%;--awb-width-medium:100%;--awb-spacing-right-medium:1.92%;--awb-spacing-left-medium:1.92%;--awb-width-small:100%;--awb-spacing-right-small:1.92%;--awb-spacing-left-small:1.92%;\"><div class=\"fusion-column-wrapper fusion-flex-justify-content-flex-start fusion-content-layout-column\"><link rel=\"preconnect\" href=\"https:\/\/fonts.googleapis.com\"><link rel=\"preconnect\" href=\"https:\/\/fonts.gstatic.com\" crossorigin><link href=\"https:\/\/fonts.googleapis.com\/css2?family=Space+Grotesk:wght@500;600;700&family=Hanken+Grotesk:wght@400;500;600;700&family=JetBrains+Mono:wght@400;500&display=swap\" rel=\"stylesheet\">\n<style>\n.fusion-page-title-bar{display:none!important;}\n.deeper-page{--paper:#ffffff;\n    --panel:#f4f8f9;\n    --panel-2:#eaf4f5;\n    --ink:#16222b;\n    --ink-soft:#46585f;\n    --muted:#8a9aa1;\n    --line:#e2eaec;\n    --line-strong:#cfdcde;\n    --teal:#1d919e;\n    --teal-d:#15707a;\n    --teal-soft:#e7f3f4;\n    --teal-glow:#6fd2dd;\n    --red:#7f191a;\n    --red-soft:#f6ebeb;\n    --navy:#0e1726;\n    --navy-2:#16243a;\n    --sans:\"Hanken Grotesk\",system-ui,sans-serif;\n    --disp:\"Space Grotesk\",system-ui,sans-serif;\n    --mono:\"JetBrains Mono\",ui-monospace,monospace;\n    --wrap:1240px;\n    color-scheme:light;}\n.deeper-page *{box-sizing:border-box}\n.deeper-page{scroll-behavior:smooth}\n.deeper-page{margin:0;background:var(--paper);color:var(--ink);font-family:var(--sans);font-size:18px;line-height:1.6;-webkit-font-smoothing:antialiased;overflow-x:hidden}\n.deeper-page h1, .deeper-page h2, .deeper-page h3, .deeper-page h4{font-family:var(--disp);font-weight:700;line-height:1.05;letter-spacing:-.02em;margin:0}\n.deeper-page p{margin:0}\n.deeper-page a{color:inherit;text-decoration:none}\n.deeper-page img{max-width:100%;height:auto;display:block}\n.deeper-page .wrap{max-width:var(--wrap);margin:0 auto;padding:0 32px}\n.deeper-page .teal{color:var(--teal)}\n.deeper-page .redc{color:var(--red)}\n.deeper-page .eyebrow{font-family:var(--mono);font-size:14px;font-weight:500;letter-spacing:.18em;text-transform:uppercase;color:var(--teal);display:inline-flex;align-items:center;gap:12px}\n.deeper-page .eyebrow::before{content:\"\";width:28px;height:2px;background:var(--teal)}\n.deeper-page .eyebrow.red{color:var(--red)}\n.deeper-page .eyebrow.red::before{background:var(--red)}\n.deeper-page .eyebrow.light{color:var(--teal-glow)}\n.deeper-page .eyebrow.light::before{background:var(--teal-glow)}\n.deeper-page .btn{display:inline-flex;align-items:center;gap:10px;font-family:var(--disp);font-weight:600;font-size:16px;padding:15px 26px;border-radius:10px;border:1.5px solid transparent;cursor:pointer;transition:transform .15s,background .15s,box-shadow .15s,border-color .15s,color .15s}\n.deeper-page .btn-primary{background:var(--teal);color:#fff;box-shadow:0 8px 22px rgba(29,145,158,.28)}\n.deeper-page .btn-primary:hover{transform:translateY(-2px);background:var(--teal-d)}\n.deeper-page .btn-ghost{background:transparent;color:var(--ink);border-color:var(--line-strong)}\n.deeper-page .btn-ghost:hover{border-color:var(--teal);color:var(--teal-d)}\n.deeper-page .btn-ghost.on-dark{color:#fff;border-color:rgba(255,255,255,.3)}\n.deeper-page .btn-ghost.on-dark:hover{border-color:var(--teal-glow);color:var(--teal-glow)}\n.deeper-page .btn-arrow::after{content:\"\u2192\"}\n.deeper-page section{position:relative}\n.deeper-page .sec{padding:108px 0}\n.deeper-page .sec-head{max-width:780px;margin-bottom:54px}\n.deeper-page .sec-head h2{font-size:clamp(34px,4.4vw,56px);margin:16px 0 0}\n.deeper-page .sec-head p{font-size:21px;color:var(--ink-soft);margin-top:20px}\n.deeper-page .reveal{opacity:0;transform:translateY(26px);transition:opacity .7s cubic-bezier(.2,.7,.2,1),transform .7s cubic-bezier(.2,.7,.2,1)}\n.deeper-page .reveal.in{opacity:1;transform:none}\n.deeper-page \/* ===== HERO ===== *\/\n  .hero{padding:150px 0 96px;overflow:hidden;background:linear-gradient(180deg,var(--panel),#fff 85%)}\n.deeper-page #neural{position:absolute;inset:0;width:100%;height:100%;z-index:0}\n.deeper-page .crumbs{position:relative;z-index:2;font-family:var(--mono);font-size:13px;color:var(--muted);margin-bottom:38px;letter-spacing:.02em}\n.deeper-page .crumbs a{color:var(--ink-soft)}\n.deeper-page .crumbs a:hover{color:var(--teal-d)}\n.deeper-page .crumbs span{margin:0 8px;color:var(--line-strong)}\n.deeper-page .hero .grid{position:relative;z-index:2;display:grid;grid-template-columns:1fr 1.08fr;gap:56px;align-items:center}\n.deeper-page .hero-logo{height:58px;width:auto;margin-bottom:26px}\n.deeper-page .hero h1{font-size:clamp(40px,5.4vw,72px);letter-spacing:-.03em;margin-top:18px}\n.deeper-page .hero .sub{font-size:clamp(18px,2vw,22px);color:var(--ink-soft);margin-top:26px;max-width:40ch}\n.deeper-page .hero-cta{display:flex;gap:16px;flex-wrap:wrap;margin-top:36px}\n.deeper-page .hero-meta{display:flex;gap:36px;margin-top:46px;flex-wrap:wrap}\n.deeper-page .hero-meta .m{display:flex;flex-direction:column}\n.deeper-page .hero-meta .mv{font-family:var(--disp);font-weight:700;font-size:30px;color:var(--ink);line-height:1}\n.deeper-page .hero-meta .ml{font-family:var(--mono);font-size:12.5px;letter-spacing:.06em;text-transform:uppercase;color:var(--muted);margin-top:8px}\n.deeper-page .hero-visual{position:relative}\n.deeper-page .shot{border-radius:16px;overflow:hidden;border:3px solid #fff;outline:1px solid var(--line);box-shadow:0 30px 70px rgba(16,34,43,.18);background:#fff}\n.deeper-page .shot img{width:100%;display:block}\n.deeper-page .hero-badge{position:absolute;left:-22px;bottom:-26px;background:#fff;border:1px solid var(--line);border-radius:14px;padding:16px 20px;box-shadow:0 18px 40px rgba(16,34,43,.16);display:flex;align-items:center;gap:14px}\n.deeper-page .hero-badge .dot{width:12px;height:12px;border-radius:50%;background:var(--red);box-shadow:0 0 0 5px var(--red-soft);flex:none}\n.deeper-page .hero-badge .t{font-family:var(--disp);font-weight:700;font-size:16px;line-height:1.2}\n.deeper-page .hero-badge .d{font-family:var(--mono);font-size:12px;color:var(--muted)}\n.deeper-page \/* ===== SUCCESSION ===== *\/\n  .succ{border-top:1px solid var(--line);border-bottom:1px solid var(--line);background:#fff}\n.deeper-page .succ .wrap{display:flex;align-items:center;gap:22px;padding-top:22px;padding-bottom:22px;flex-wrap:wrap}\n.deeper-page .succ .tag{font-family:var(--mono);font-size:12.5px;letter-spacing:.12em;text-transform:uppercase;color:#fff;background:var(--red);border-radius:5px;padding:6px 10px;flex:none}\n.deeper-page .succ p{font-size:16.5px;color:var(--ink-soft);flex:1;min-width:260px}\n.deeper-page .succ p b{color:var(--ink)}\n.deeper-page .succ s{color:var(--muted)}\n.deeper-page \/* ===== POURQUOI ===== *\/\n  .why{display:grid;grid-template-columns:repeat(3,1fr);gap:22px}\n.deeper-page .why-c{border:1px solid var(--line);border-radius:18px;padding:32px 30px;background:#fff;transition:transform .2s,box-shadow .2s}\n.deeper-page .why-c:hover{transform:translateY(-4px);box-shadow:0 24px 50px rgba(16,34,43,.08)}\n.deeper-page .why-c .n{font-family:var(--mono);font-size:13px;letter-spacing:.14em;color:var(--teal);font-weight:600}\n.deeper-page .why-c:nth-child(2) .n{color:var(--red)}\n.deeper-page .why-c h3{font-size:24px;margin:14px 0 12px}\n.deeper-page .why-c p{font-size:16.5px;color:var(--ink-soft);line-height:1.55}\n.deeper-page \/* ===== BOITE A OUTILS ===== *\/\n  .tools-sec{background:var(--panel)}\n.deeper-page .tools{display:grid;grid-template-columns:repeat(5,1fr);gap:16px}\n.deeper-page .tool{background:#fff;border:1px solid var(--line);border-radius:16px;padding:24px 22px;display:flex;flex-direction:column;gap:12px;transition:border-color .2s,transform .2s,box-shadow .2s}\n.deeper-page .tool:hover{border-color:var(--teal);transform:translateY(-3px);box-shadow:0 18px 40px rgba(16,34,43,.08)}\n.deeper-page .tool .ic{width:44px;height:44px;border-radius:11px;background:var(--teal-soft);color:var(--teal-d);display:flex;align-items:center;justify-content:center}\n.deeper-page .tool:nth-child(even) .ic{background:var(--red-soft);color:var(--red)}\n.deeper-page .tool .ic svg{width:24px;height:24px;fill:none;stroke:currentColor;stroke-width:1.8;stroke-linecap:round;stroke-linejoin:round}\n.deeper-page .tool h3{font-size:19px;line-height:1.15}\n.deeper-page .tool p{font-size:14.5px;color:var(--ink-soft);line-height:1.5}\n.deeper-page .tool .k{margin-top:auto;font-family:var(--mono);font-size:11.5px;color:var(--muted);line-height:1.5;padding-top:10px;border-top:1px dashed var(--line)}\n.deeper-page .tool.feat{background:var(--navy);border-color:var(--navy);color:#fff}\n.deeper-page .tool.feat .ic{background:rgba(111,210,221,.15);color:var(--teal-glow)}\n.deeper-page .tool.feat p{color:#b6c4d0}\n.deeper-page .tool.feat .k{color:#7d8e9c;border-top-color:rgba(255,255,255,.14)}\n.deeper-page .tools-note{margin-top:30px;font-size:16px;color:var(--ink-soft);max-width:80ch}\n.deeper-page .tools-note b{color:var(--ink)}\n.deeper-page \/* ===== PIPELINE ===== *\/\n  .pipe{display:grid;grid-template-columns:1.25fr 1fr;gap:60px;align-items:center}\n.deeper-page .pipe h2{font-size:clamp(32px,4vw,50px);margin-top:16px}\n.deeper-page .pipe .lead{font-size:19px;color:var(--ink-soft);margin-top:22px}\n.deeper-page .pts{margin-top:28px;display:flex;flex-direction:column;gap:16px}\n.deeper-page .pt{display:grid;grid-template-columns:28px 1fr;gap:14px;align-items:start;font-size:16.5px;color:var(--ink-soft);line-height:1.5}\n.deeper-page .pt::before{content:\"\u2713\";color:var(--teal);font-weight:800;font-size:18px;line-height:1.4}\n.deeper-page .pt b{color:var(--ink)}\n.deeper-page \/* ===== STATS ===== *\/\n  .stats{background:var(--navy);color:#fff;padding:84px 0}\n.deeper-page .stats .grid{display:grid;grid-template-columns:repeat(4,1fr);gap:40px}\n.deeper-page .stat{border-top:2px solid rgba(255,255,255,.12);padding-top:22px}\n.deeper-page .stat:nth-child(2){border-top-color:var(--teal)}\n.deeper-page .stat:nth-child(4){border-top-color:var(--red)}\n.deeper-page .stat .n{font-family:var(--disp);font-weight:700;font-size:clamp(40px,4.6vw,60px);line-height:.95;letter-spacing:-.03em}\n.deeper-page .stat .n .u{font-size:.45em;color:var(--teal-glow);font-weight:600;margin-left:6px}\n.deeper-page .stat .l{font-size:16.5px;color:#aebccb;margin-top:12px;line-height:1.4;max-width:24ch}\n.deeper-page \/* ===== METHODE ===== *\/\n  .steps{display:grid;grid-template-columns:repeat(3,1fr);gap:0;border:1px solid var(--line);border-radius:20px;overflow:hidden;background:#fff}\n.deeper-page .step{padding:34px 32px;border-right:1px solid var(--line);border-bottom:1px solid var(--line);position:relative}\n.deeper-page .step:nth-child(3n){border-right:none}\n.deeper-page .step:nth-child(n+4){border-bottom:none}\n.deeper-page .step .sn{font-family:var(--disp);font-weight:700;font-size:44px;color:var(--teal-soft);line-height:1;-webkit-text-stroke:1.5px var(--teal)}\n.deeper-page .step:nth-child(even) .sn{color:var(--red-soft);-webkit-text-stroke-color:var(--red)}\n.deeper-page .step .mod{font-family:var(--mono);font-size:12px;letter-spacing:.1em;text-transform:uppercase;color:var(--muted);margin-top:18px}\n.deeper-page .step h3{font-size:22px;margin:6px 0 10px}\n.deeper-page .step p{font-size:16px;color:var(--ink-soft);line-height:1.55}\n.deeper-page \/* ===== DUO (VLM \/ reflets) ===== *\/\n  .duo{display:grid;grid-template-columns:1fr 1fr;gap:28px}\n.deeper-page .spot{border-radius:20px;overflow:hidden;border:1px solid var(--line);background:#fff;display:flex;flex-direction:column}\n.deeper-page .spot .img{aspect-ratio:16\/9;overflow:hidden;background:#0b0f14;position:relative}\n.deeper-page .spot .img img{width:100%;height:100%;object-fit:cover}\n.deeper-page .spot .img .lbl{position:absolute;left:16px;top:16px;font-family:var(--mono);font-size:12px;letter-spacing:.08em;text-transform:uppercase;background:rgba(14,23,38,.82);color:var(--teal-glow);padding:6px 10px;border-radius:6px;backdrop-filter:blur(4px)}\n.deeper-page .spot .body{padding:32px 32px 34px;display:flex;flex-direction:column;gap:14px;flex:1}\n.deeper-page .spot h3{font-size:clamp(24px,2.4vw,30px)}\n.deeper-page .spot p{font-size:16.5px;color:var(--ink-soft);line-height:1.55}\n.deeper-page .spot .chips{display:flex;flex-wrap:wrap;gap:8px;margin-top:auto;padding-top:8px}\n.deeper-page .chips span{font-family:var(--mono);font-size:12.5px;color:var(--ink-soft);background:var(--panel);border:1px solid var(--line);border-radius:5px;padding:6px 11px}\n.deeper-page \/* ===== SPECS ===== *\/\n  .specs-sec{background:linear-gradient(180deg,#fff,var(--panel))}\n.deeper-page .specs{display:grid;grid-template-columns:1fr 1fr;gap:0 56px}\n.deeper-page .spec{display:grid;grid-template-columns:200px 1fr;gap:20px;padding:20px 0;border-bottom:1px solid var(--line-strong)}\n.deeper-page .spec .k{font-family:var(--mono);font-size:13px;letter-spacing:.08em;text-transform:uppercase;color:var(--teal-d);padding-top:3px}\n.deeper-page .spec:nth-child(4n+3) .k, .deeper-page .spec:nth-child(4n+4) .k{color:var(--red)}\n.deeper-page .spec .v{font-size:16.5px;color:var(--ink);line-height:1.5}\n.deeper-page .spec .v small{display:block;color:var(--muted);font-size:14px;margin-top:2px}\n.deeper-page \/* ===== APPLICATIONS ===== *\/\n  .cases{display:grid;grid-template-columns:repeat(3,1fr);gap:22px}\n.deeper-page .case{border-radius:18px;overflow:hidden;border:1px solid var(--line);background:#fff;display:flex;flex-direction:column;transition:transform .2s,box-shadow .2s}\n.deeper-page .case:hover{transform:translateY(-4px);box-shadow:0 24px 50px rgba(16,34,43,.1)}\n.deeper-page .case .ci{aspect-ratio:560\/340;overflow:hidden;background:#0b0f14}\n.deeper-page .case .ci img{width:100%;height:100%;object-fit:cover;transition:transform .5s}\n.deeper-page .case:hover .ci img{transform:scale(1.05)}\n.deeper-page .case .cb{padding:22px 24px 24px;display:flex;flex-direction:column;gap:8px;flex:1}\n.deeper-page .case .sector{font-family:var(--mono);font-size:12px;letter-spacing:.08em;text-transform:uppercase;color:var(--teal-d)}\n.deeper-page .case:nth-child(even) .sector{color:var(--red)}\n.deeper-page .case h3{font-size:21px;line-height:1.15}\n.deeper-page .case p{font-size:15px;color:var(--ink-soft);line-height:1.5}\n.deeper-page .case .kpis{display:flex;flex-wrap:wrap;gap:6px;margin-top:auto;padding-top:10px}\n.deeper-page .case .kpis span{font-family:var(--mono);font-size:12px;color:var(--ink);background:var(--teal-soft);border-radius:5px;padding:5px 9px}\n.deeper-page .case:nth-child(even) .kpis span{background:var(--red-soft)}\n.deeper-page .cases-more{margin-top:34px;display:flex;justify-content:space-between;align-items:center;gap:20px;flex-wrap:wrap;font-size:16.5px;color:var(--ink-soft)}\n.deeper-page \/* ===== OFFRES ===== *\/\n  .offres-sec{background:var(--navy);color:#fff;overflow:hidden}\n.deeper-page #neural2{position:absolute;inset:0;width:100%;height:100%;opacity:.35}\n.deeper-page .offres-sec .wrap{position:relative;z-index:2}\n.deeper-page .offres-sec .sec-head h2{color:#fff}\n.deeper-page .offres-sec .sec-head p{color:#aebccb}\n.deeper-page .offres{display:grid;grid-template-columns:1fr 1fr;gap:24px}\n.deeper-page .offre{border:1px solid rgba(255,255,255,.14);border-radius:20px;padding:38px 36px;background:rgba(255,255,255,.03);display:flex;flex-direction:column;gap:16px}\n.deeper-page .offre.main{background:#fff;color:var(--ink);border-color:#fff}\n.deeper-page .offre .ot{font-family:var(--mono);font-size:12.5px;letter-spacing:.14em;text-transform:uppercase;color:var(--teal-glow)}\n.deeper-page .offre.main .ot{color:var(--teal-d)}\n.deeper-page .offre h3{font-size:28px}\n.deeper-page .offre p{font-size:16.5px;color:#b6c4d0;line-height:1.55}\n.deeper-page .offre.main p{color:var(--ink-soft)}\n.deeper-page .offre ul{list-style:none;padding:0;margin:4px 0 0;display:flex;flex-direction:column;gap:10px}\n.deeper-page .offre li{display:grid;grid-template-columns:22px 1fr;gap:10px;font-size:16px;line-height:1.45;color:#dbe4ec}\n.deeper-page .offre.main li{color:var(--ink-soft)}\n.deeper-page .offre li::before{content:\"\u2192\";color:var(--teal-glow);font-weight:700}\n.deeper-page .offre.main li::before{color:var(--teal)}\n.deeper-page .offre .btn{align-self:flex-start;margin-top:auto}\n.deeper-page .cir{margin-top:30px;display:flex;align-items:center;gap:14px;font-size:15.5px;color:#aebccb;flex-wrap:wrap}\n.deeper-page .cir b{color:#fff}\n.deeper-page .cir .pill{font-family:var(--mono);font-size:12px;letter-spacing:.1em;text-transform:uppercase;border:1px solid rgba(111,210,221,.5);color:var(--teal-glow);border-radius:999px;padding:6px 14px}\n.deeper-page \/* ===== FAQ ===== *\/\n  .faq-grid{display:grid;grid-template-columns:.8fr 1.2fr;gap:64px;align-items:start}\n.deeper-page .faq-grid .sec-head{position:sticky;top:110px;margin-bottom:0}\n.deeper-page .faq details{border-bottom:1px solid var(--line-strong);padding:4px 0}\n.deeper-page .faq details:first-child{border-top:1px solid var(--line-strong)}\n.deeper-page .faq summary{list-style:none;cursor:pointer;display:flex;justify-content:space-between;gap:20px;align-items:center;padding:20px 0;font-family:var(--disp);font-weight:600;font-size:19px;color:var(--ink)}\n.deeper-page .faq summary::-webkit-details-marker{display:none}\n.deeper-page .faq summary::after{content:\"+\";font-family:var(--mono);font-size:22px;color:var(--teal);transition:transform .2s;flex:none}\n.deeper-page .faq details[open] summary::after{transform:rotate(45deg)}\n.deeper-page .faq details p{font-size:16.5px;color:var(--ink-soft);padding:0 40px 22px 0;line-height:1.6}\n.deeper-page \/* ===== CONTACT ===== *\/\n  .contact{background:linear-gradient(160deg,var(--navy),#0a1018);color:#fff;border-radius:26px;padding:64px;display:grid;grid-template-columns:1fr 1fr;gap:60px;position:relative;overflow:hidden}\n.deeper-page #neural3{position:absolute;inset:0;width:100%;height:100%;opacity:.5}\n.deeper-page .contact>*{position:relative;z-index:2}\n.deeper-page .contact h2{font-size:clamp(32px,4vw,48px);color:#fff;max-width:16ch;margin-top:16px}\n.deeper-page .contact .lead{font-size:19px;color:#aebccb;margin-top:20px;max-width:42ch}\n.deeper-page .cinfo{margin-top:34px}\n.deeper-page .ci2{padding:18px 0;border-bottom:1px solid rgba(255,255,255,.12);display:flex;gap:16px;align-items:center}\n.deeper-page .ci2:first-child{border-top:1px solid rgba(255,255,255,.12)}\n.deeper-page .ci2 .k{font-family:var(--mono);font-size:12px;letter-spacing:.14em;text-transform:uppercase;color:var(--teal-glow);min-width:96px}\n.deeper-page .ci2 .v{font-size:18px;color:#fff}\n.deeper-page .form{background:#fff;border-radius:18px;padding:34px;color:var(--ink)}\n.deeper-page .form h3{font-size:22px;margin-bottom:6px}\n.deeper-page .form p.fh{font-size:14.5px;color:var(--muted);margin-bottom:22px}\n.deeper-page .field{margin-bottom:16px}\n.deeper-page .field label{display:block;font-size:13px;font-weight:600;color:var(--ink-soft);margin-bottom:7px}\n.deeper-page .field input, .deeper-page .field textarea{width:100%;border:1.5px solid var(--line-strong);border-radius:10px;padding:13px 15px;font-family:var(--sans);font-size:15px;color:var(--ink);background:var(--panel);transition:border-color .15s,background .15s}\n.deeper-page .field input:focus, .deeper-page .field textarea:focus{outline:none;border-color:var(--teal);background:#fff}\n.deeper-page .field textarea{resize:vertical;min-height:110px}\n.deeper-page .form .btn-primary{width:100%;justify-content:center;margin-top:6px}\n.deeper-page .form .note{font-size:12.5px;color:var(--muted);margin-top:14px;text-align:center}\n@media(max-width:1180px){\n  .deeper-page .tools{grid-template-columns:repeat(3,1fr)}\n}\n@media(max-width:980px){\n  .deeper-page .hero .grid{grid-template-columns:1fr;gap:56px}\n  .deeper-page .why{grid-template-columns:1fr}\n  .deeper-page .pipe{grid-template-columns:1fr;gap:40px}\n  .deeper-page .stats .grid{grid-template-columns:1fr 1fr}\n  .deeper-page .steps{grid-template-columns:1fr 1fr}\n  .deeper-page .step{border-right:1px solid var(--line)!important;border-bottom:1px solid var(--line)!important}\n  .deeper-page .step:nth-child(2n){border-right:none!important}\n  .deeper-page .step:nth-child(n+5){border-bottom:none!important}\n  .deeper-page .duo{grid-template-columns:1fr}\n  .deeper-page .specs{grid-template-columns:1fr}\n  .deeper-page .cases{grid-template-columns:1fr 1fr}\n  .deeper-page .offres{grid-template-columns:1fr}\n  .deeper-page .faq-grid{grid-template-columns:1fr;gap:36px}\n  .deeper-page .faq-grid .sec-head{position:static}\n  .deeper-page .contact{grid-template-columns:1fr;padding:44px 32px}\n}\n@media(max-width:640px){\n  .deeper-page{font-size:17px}\n  .deeper-page .wrap{padding:0 16px}\n  .deeper-page .sec{padding:76px 0}\n  .deeper-page .hero{padding:118px 0 80px}\n  .deeper-page .hero-badge{position:relative;left:auto;bottom:auto;margin:-18px 0 0 12px;width:fit-content;padding:12px 16px}\n  .deeper-page .tools{grid-template-columns:1fr 1fr}\n  .deeper-page .stats .grid{grid-template-columns:1fr}\n  .deeper-page .steps{grid-template-columns:1fr}\n  .deeper-page .step{border-right:none!important;border-bottom:1px solid var(--line)!important}\n  .deeper-page .step:last-child{border-bottom:none!important}\n  .deeper-page .spec{grid-template-columns:1fr;gap:6px}\n  .deeper-page .cases{grid-template-columns:1fr}\n  .deeper-page .contact{padding:36px 20px;border-radius:20px}\n  .deeper-page .form{padding:24px 20px}\n  .deeper-page .offre{padding:30px 24px}\n}\n@media(max-width:420px){\n  .deeper-page .tools{grid-template-columns:1fr}\n}\n@media (prefers-reduced-motion: reduce){\n  .deeper-page .reveal{opacity:1;transform:none;transition:none}\n  .deeper-page{scroll-behavior:auto}\n}\n\n.deeper-page #outils,.deeper-page #methode,.deeper-page #applications,.deeper-page #offres,.deeper-page #faq,.deeper-page #contact,.deeper-page #specs,.deeper-page #innovations{scroll-margin-top:100px}\n.deeper-page h1,.deeper-page h2,.deeper-page h3,.deeper-page h4{color:inherit!important}\n<\/style>\n<div class=\"deeper-page\">\n<script type=\"application\/ld+json\">\n{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"SoftwareApplication\",\"name\":\"Deeper\",\"applicationCategory\":\"BusinessApplication\",\"applicationSubCategory\":\"Vision industrielle par intelligence artificielle\",\"operatingSystem\":\"PC avec GPU (local) ou serveur web (SaaS)\",\"description\":\"Logiciel de vision industrielle par IA : d\u00e9tection d'anomalies, d\u00e9tection et segmentation, OCR et codes, mesure, comptage, couleur, polarisation et mod\u00e8les vision-langage (VLM), programmable sans code par pipelines de n\u0153uds.\",\"publisher\":{\"@type\":\"Organization\",\"name\":\"CIM Atlantique\",\"url\":\"https:\/\/www.cim-atlantique.com\/\"},\"url\":\"https:\/\/www.cim-atlantique.com\/fr\/deeper\/\"},{\"@type\":\"FAQPage\",\"mainEntity\":[{\"@type\":\"Question\",\"name\":\"Deeper remplace-t-il Deep Capture ?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Oui. Deeper est la nouvelle g\u00e9n\u00e9ration de notre logiciel de vision par IA. Il reprend l'exp\u00e9rience acquise avec Deep Capture en deep learning supervis\u00e9 et l'\u00e9largit : d\u00e9tection d'anomalies, mesure, OCR, comptage, polarisation, mod\u00e8les vision-langage, le tout programmable sans code.\"}},{\"@type\":\"Question\",\"name\":\"Faut-il \u00eatre expert en vision ou en intelligence artificielle pour utiliser Deeper ?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Non. Les contr\u00f4les se construisent sans code, en assemblant des briques dans un \u00e9diteur visuel. Nos ing\u00e9nieurs peuvent livrer l'application cl\u00e9 en main, ou former vos \u00e9quipes pour qu'elles la fassent \u00e9voluer elles-m\u00eames.\"}},{\"@type\":\"Question\",\"name\":\"Quelles cam\u00e9ras sont compatibles avec Deeper ?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Deeper fonctionne notamment avec les cam\u00e9ras industrielles Baumer, Basler et SVS-Vistek, et peut aussi recevoir des images \u00e0 distance (t\u00e9l\u00e9phone, tablette) pour la phase de mise au point.\"}},{\"@type\":\"Question\",\"name\":\"Deeper fonctionne-t-il dans le cloud ou en local ?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Les deux : en local sur un PC \u00e9quip\u00e9 d'un GPU, au plus pr\u00e8s de la ligne, ou en SaaS sur un serveur web.\"}},{\"@type\":\"Question\",\"name\":\"Comment Deeper communique-t-il avec mon automate ?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Le r\u00e9sultat du contr\u00f4le (verdict OK\/NOK, mesures, texte lu) peut \u00eatre transmis \u00e0 l'automate ou au syst\u00e8me d'information via OPC-UA, Modbus TCP ou requ\u00eate HTTP.\"}},{\"@type\":\"Question\",\"name\":\"Peut-on tester Deeper sur nos propres pi\u00e8ces ?\",\"acceptedAnswer\":{\"@type\":\"Answer\",\"text\":\"Oui. Envoyez-nous quelques images de vos pi\u00e8ces, bonnes et d\u00e9fectueuses : nous \u00e9valuons la faisabilit\u00e9 avec Deeper et vous pr\u00e9sentons les r\u00e9sultats, en visio ou dans nos ateliers.\"}}]}]}\n<\/script>\n<!-- ===== HERO ===== -->\n<section class=\"hero\" id=\"top\">\n  <canvas id=\"neural\" aria-hidden=\"true\"><\/canvas>\n  <div class=\"wrap\">\n    <nav class=\"crumbs\" aria-label=\"Fil d'Ariane\">\n      <a href=\"https:\/\/www.cim-atlantique.com\/fr\/\">Accueil<\/a><span>\/<\/span><a href=\"https:\/\/www.cim-atlantique.com\/fr\/vision-industrielle\/\">Vision industrielle<\/a><span>\/<\/span>Deeper\n    <\/nav>\n    <div class=\"grid\">\n      <div class=\"hero-copy\">\n        <img decoding=\"async\" class=\"lazyload hero-logo\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27120%27%20viewBox%3D%270%200%20560%20120%27%3E%3Crect%20width%3D%27560%27%20height%3D%27120%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/deeper-logo-light.png\" alt=\"Deeper\" title=\"\">\n        <span class=\"eyebrow\">D\u00e9velopp\u00e9 par CIM Atlantique<\/span>\n        <h1>Le logiciel de vision industrielle qui r\u00e9unit <span class=\"teal\">tous les outils de l'IA<\/span>.<\/h1>\n        <p class=\"sub\">Deeper r\u00e9unit dans une m\u00eame plateforme les meilleurs algorithmes d'intelligence artificielle et de vision \u2014 d\u00e9tection d'anomalies, OCR, mesure, comptage, mod\u00e8les vision-langage \u2014 et les assemble sans une ligne de code.<\/p>\n        <div class=\"hero-cta\">\n          <a href=\"#contact\" class=\"btn btn-primary btn-arrow\">Tester Deeper sur mes pi\u00e8ces<\/a>\n          <a href=\"#applications\" class=\"btn btn-ghost\">Voir les applications<\/a>\n        <\/div>\n        <div class=\"hero-meta\">\n          <div class=\"m\"><span class=\"mv\">+80<\/span><span class=\"ml\">briques de traitement<\/span><\/div>\n          <div class=\"m\"><span class=\"mv\">0<\/span><span class=\"ml\">ligne de code<\/span><\/div>\n          <div class=\"m\"><span class=\"mv\">+12 ans<\/span><span class=\"ml\">de vision par IA<\/span><\/div>\n        <\/div>\n      <\/div>\n      <div class=\"hero-visual\">\n        <div class=\"shot\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%271500%27%20height%3D%27686%27%20viewBox%3D%270%200%201500%20686%27%3E%3Crect%20width%3D%271500%27%20height%3D%27686%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/deeper-ui-c.jpg\" alt=\"Interface de Deeper en mode simulation : d\u00e9tection d&#039;anomalies sur des pi\u00e8ces, avec score par pi\u00e8ce\" width=\"1500\" height=\"686\" title=\"\"><\/div>\n        <div class=\"hero-badge\">\n          <span class=\"dot\"><\/span>\n          <span><span class=\"t\">D\u00e9tection d'anomalies<\/span><br><span class=\"d\">score et carte de chaleur par pi\u00e8ce<\/span><\/span>\n        <\/div>\n      <\/div>\n    <\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== SUCCESSION DEEP CAPTURE ===== -->\n<div class=\"succ\">\n  <div class=\"wrap\">\n    <span class=\"tag\">Nouvelle g\u00e9n\u00e9ration<\/span>\n    <p><b>Deeper succ\u00e8de \u00e0 <s>Deep Capture<\/s>.<\/b> Le deep learning supervis\u00e9 qui a fait ses preuves sur nos lignes clientes reste au c\u0153ur de l'outil, d\u00e9sormais entour\u00e9 d'une bo\u00eete \u00e0 outils compl\u00e8te et d'un \u00e9diteur no-code.<\/p>\n  <\/div>\n<\/div>\n\n<!-- ===== POURQUOI ===== -->\n<section class=\"sec\" id=\"pourquoi\">\n  <div class=\"wrap\">\n    <div class=\"sec-head reveal\">\n      <span class=\"eyebrow\">Pourquoi Deeper<\/span>\n      <h2>Le bon algorithme pour chaque contr\u00f4le.<\/h2>\n      <p>Aucun algorithme ne sait tout faire. Un d\u00e9faut d'aspect, une cote, un marquage ou un comptage demandent des outils diff\u00e9rents : Deeper les r\u00e9unit et vous laisse les combiner.<\/p>\n    <\/div>\n    <div class=\"why\">\n      <div class=\"why-c reveal\">\n        <div class=\"n\">01 \u2014 BO\u00ceTE \u00c0 OUTILS<\/div>\n        <h3>Toutes les approches, au m\u00eame endroit<\/h3>\n        <p>Deep learning, d\u00e9tection d'anomalies, vision \u00ab classique \u00bb \u00e0 r\u00e8gles, OCR, polarisation, VLM : on choisit la technique adapt\u00e9e au d\u00e9faut, pas l'inverse.<\/p>\n      <\/div>\n      <div class=\"why-c reveal\">\n        <div class=\"n\">02 \u2014 NO-CODE<\/div>\n        <h3>Une application se construit, elle ne se code pas<\/h3>\n        <p>Chaque contr\u00f4le est un pipeline visuel : on relie des briques, on teste sur ses images, on ajuste les param\u00e8tres en direct.<\/p>\n      <\/div>\n      <div class=\"why-c reveal\">\n        <div class=\"n\">03 \u2014 INDUSTRIEL<\/div>\n        <h3>Pens\u00e9 pour la ligne de production<\/h3>\n        <p>Calcul sur GPU pour tenir la cadence, multicam\u00e9ra, \u00e9cran op\u00e9rateur personnalisable et dialogue direct avec l'automate.<\/p>\n      <\/div>\n    <\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== BOITE A OUTILS ===== -->\n<section class=\"sec tools-sec\" id=\"outils\">\n  <div class=\"wrap\">\n    <div class=\"sec-head reveal\">\n      <span class=\"eyebrow red\">La bo\u00eete \u00e0 outils<\/span>\n      <h2>Plus de 80 briques, dix familles d'outils.<\/h2>\n      <p>Des mod\u00e8les d'IA de derni\u00e8re g\u00e9n\u00e9ration aux outils de vision \u00e9prouv\u00e9s, chaque brique s'ins\u00e8re dans un pipeline et se combine avec les autres.<\/p>\n    <\/div>\n    <div class=\"tools\">\n      <div class=\"tool reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><path d=\"M12 3l9 5v8l-9 5-9-5V8z\"\/><path d=\"M12 12l9-4M12 12v9M12 12L3 8\"\/><\/svg><\/span>\n        <h3>D\u00e9tection d'anomalies<\/h3>\n        <p>Apprend l'aspect d'une pi\u00e8ce bonne et signale tout \u00e9cart, m\u00eame un d\u00e9faut jamais vu, avec un score et une carte de chaleur.<\/p>\n        <div class=\"k\">anomalib \u00b7 AnomalyDINO \u00b7 RareCLIP<\/div>\n      <\/div>\n      <div class=\"tool reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><rect x=\"3\" y=\"3\" width=\"7\" height=\"7\" rx=\"1\"\/><rect x=\"14\" y=\"14\" width=\"7\" height=\"7\" rx=\"1\"\/><path d=\"M14 3h7v7M3 14v7h7\"\/><\/svg><\/span>\n        <h3>D\u00e9tection & classification<\/h3>\n        <p>Localise, classe et trie objets et d\u00e9fauts en temps r\u00e9el, y compris par simple description textuelle de l'objet recherch\u00e9.<\/p>\n        <div class=\"k\">YOLO \u00b7 YOLO-World \u00b7 Grounding DINO<\/div>\n      <\/div>\n      <div class=\"tool reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><path d=\"M4 18c3-8 6-11 8-11s5 3 8 11\"\/><path d=\"M4 18h16\"\/><circle cx=\"12\" cy=\"7\" r=\"2\"\/><\/svg><\/span>\n        <h3>Segmentation<\/h3>\n        <p>D\u00e9toure pr\u00e9cis\u00e9ment chaque pi\u00e8ce ou zone pour mesurer une surface, isoler une r\u00e9gion ou redresser une \u00e9tiquette.<\/p>\n        <div class=\"k\">SAM 2 \u00b7 SAM 3 \u00b7 FastSAM<\/div>\n      <\/div>\n      <div class=\"tool reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><path d=\"M4 6V4h16v2M9 20h6M12 4v16\"\/><\/svg><\/span>\n        <h3>OCR & codes<\/h3>\n        <p>Relit dates, lots et marquages, m\u00eame jet d'encre irr\u00e9gulier ou texte inclin\u00e9. Lecture des codes-barres, QR et Datamatrix.<\/p>\n        <div class=\"k\">PaddleOCR \u00b7 OpenOCR \u00b7 OCR par VLM \u00b7 codes<\/div>\n      <\/div>\n      <div class=\"tool reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><path d=\"M3 17L17 3l4 4L7 21H3z\"\/><path d=\"M7 13l2 2M10 10l2 2M13 7l2 2\"\/><\/svg><\/span>\n        <h3>Mesure & g\u00e9om\u00e9trie<\/h3>\n        <p>Mesure d'ar\u00eates sub-pixel, ajustement de droites et de cercles, distances, angles, recalage de pi\u00e8ce et recherche de motif.<\/p>\n        <div class=\"k\">caliper \u00b7 fit \u00b7 fixture \u00b7 pattern match<\/div>\n      <\/div>\n      <div class=\"tool reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><circle cx=\"9\" cy=\"9\" r=\"2\"\/><circle cx=\"15\" cy=\"9\" r=\"2\"\/><circle cx=\"9\" cy=\"15\" r=\"2\"\/><circle cx=\"15\" cy=\"15\" r=\"2\"\/><circle cx=\"12\" cy=\"12\" r=\"10\"\/><\/svg><\/span>\n        <h3>Comptage<\/h3>\n        <p>Compte des objets \u00e0 partir de quelques exemples ou d'une description, y compris des produits empil\u00e9s.<\/p>\n        <div class=\"k\">CountGD \u00b7 GECO2 \u00b7 LOCA \u00b7 FamNet<\/div>\n      <\/div>\n      <div class=\"tool reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><circle cx=\"12\" cy=\"12\" r=\"9\"\/><path d=\"M12 3a9 9 0 0 1 0 18z\"\/><\/svg><\/span>\n        <h3>Couleur<\/h3>\n        <p>V\u00e9rifie la teinte d'une zone par rapport \u00e0 une r\u00e9f\u00e9rence et analyse la r\u00e9partition des couleurs.<\/p>\n        <div class=\"k\">color check \u00b7 histogrammes<\/div>\n      <\/div>\n      <div class=\"tool reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><path d=\"M12 3v18M3 12h18\"\/><circle cx=\"12\" cy=\"12\" r=\"5\"\/><\/svg><\/span>\n        <h3>Polarisation & retraitement<\/h3>\n        <p>Supprime les reflets, r\u00e9v\u00e8le les micro-d\u00e9fauts et pr\u00e9pare l'image : contraste, filtres, redressement, d\u00e9coupe.<\/p>\n        <div class=\"k\">Stokes \u00b7 anti-reflet \u00b7 deskew \u00b7 +20 filtres<\/div>\n      <\/div>\n      <div class=\"tool feat reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><path d=\"M4 5h16v10H9l-5 4z\"\/><path d=\"M8 9h8M8 12h5\"\/><\/svg><\/span>\n        <h3>VLM \u2014 vision par prompt<\/h3>\n        <p>Des mod\u00e8les vision-langage qui lisent, d\u00e9crivent et r\u00e9pondent \u00e0 une question sur l'image. Id\u00e9al pour d\u00e9marrer vite un contr\u00f4le.<\/p>\n        <div class=\"k\">OCR \u00b7 d\u00e9tection \u00b7 description \u00b7 question\/r\u00e9ponse<\/div>\n      <\/div>\n      <div class=\"tool reveal\">\n        <span class=\"ic\"><svg viewBox=\"0 0 24 24\"><path d=\"M5 12h4l2-5 3 10 2-5h3\"\/><\/svg><\/span>\n        <h3>D\u00e9cision & communication<\/h3>\n        <p>Seuils, logique ET\/OU, verdict final OK\/NOK, puis envoi du r\u00e9sultat \u00e0 l'automate ou au syst\u00e8me d'information.<\/p>\n        <div class=\"k\">OPC-UA \u00b7 Modbus TCP \u00b7 HTTP<\/div>\n      <\/div>\n    <\/div>\n    <p class=\"tools-note reveal\"><b>Le bon outil, pas un outil unique.<\/b> Sur une m\u00eame pi\u00e8ce, Deeper peut encha\u00eener un redressement, une d\u00e9tection d'anomalie, une lecture de lot et une mesure, puis rendre un verdict unique \u00e0 la machine.<\/p>\n  <\/div>\n<\/section>\n\n<!-- ===== PIPELINE NO-CODE ===== -->\n<section class=\"sec\" id=\"pipeline\">\n  <div class=\"wrap pipe\">\n    <div class=\"reveal\">\n      <div class=\"shot\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%271500%27%20height%3D%27683%27%20viewBox%3D%270%200%201500%20683%27%3E%3Crect%20width%3D%271500%27%20height%3D%27683%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/deeper-pipeline-c.jpg\" alt=\"\u00c9diteur de pipeline de Deeper : n\u0153uds de polarisation, segmentation, redressement et VLM reli\u00e9s entre eux\" width=\"1500\" height=\"683\" loading=\"lazy\" title=\"\"><\/div>\n    <\/div>\n    <div class=\"reveal\">\n      <span class=\"eyebrow\">Programmation no-code<\/span>\n      <h2>Un contr\u00f4le = un pipeline visuel.<\/h2>\n      <p class=\"lead\">Dans l'\u00e9diteur de Deeper, on fait glisser les briques sur le plan de travail et on les relie : cam\u00e9ra, pr\u00e9traitement, IA, mesure, d\u00e9cision, sortie automate.<\/p>\n      <div class=\"pts\">\n        <div class=\"pt\"><span><b>Lisible par tous.<\/b> Le pipeline montre exactement ce que fait le contr\u00f4le, \u00e9tape par \u00e9tape, avec les images interm\u00e9diaires.<\/span><\/div>\n        <div class=\"pt\"><span><b>R\u00e9glable en production.<\/b> Seuils, zones d'int\u00e9r\u00eat et param\u00e8tres cam\u00e9ra s'ajustent sans red\u00e9velopper l'application.<\/span><\/div>\n        <div class=\"pt\"><span><b>R\u00e9utilisable.<\/b> Chaque pipeline s'enregistre comme un preset : changement de r\u00e9f\u00e9rence en un clic.<\/span><\/div>\n        <div class=\"pt\"><span><b>\u00c9volutif.<\/b> Un nouveau d\u00e9faut, une nouvelle r\u00e9f\u00e9rence ? On ajoute une brique au lieu de tout reprendre.<\/span><\/div>\n      <\/div>\n    <\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== STATS ===== -->\n<section class=\"stats\" aria-label=\"Deeper en chiffres\">\n  <div class=\"wrap\">\n    <div class=\"grid\">\n      <div class=\"stat reveal\"><div class=\"n\" data-count=\"12\" data-prefix=\"+\">+12<span class=\"u\">ans<\/span><\/div><div class=\"l\">d'expertise en vision par IA et deep learning<\/div><\/div>\n      <div class=\"stat reveal\"><div class=\"n\">< 0,1<span class=\"u\">mm<\/span><\/div><div class=\"l\">d\u00e9fauts d\u00e9tect\u00e9s sur pi\u00e8ces laqu\u00e9es<\/div><\/div>\n      <div class=\"stat reveal\"><div class=\"n\" data-count=\"120\">120<span class=\"u\">\/min<\/span><\/div><div class=\"l\">flacons inspect\u00e9s sur 8 cam\u00e9ras<\/div><\/div>\n      <div class=\"stat reveal\"><div class=\"n\" data-count=\"80\">80<span class=\"u\">\/min<\/span><\/div><div class=\"l\">fromages contr\u00f4l\u00e9s, multi-d\u00e9fauts<\/div><\/div>\n    <\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== METHODE ===== -->\n<section class=\"sec\" id=\"methode\">\n  <div class=\"wrap\">\n    <div class=\"sec-head reveal\">\n      <span class=\"eyebrow\">De l'image \u00e0 la ligne<\/span>\n      <h2>Six \u00e9tapes, une seule plateforme.<\/h2>\n      <p>Acquisition, entra\u00eenement, mise au point et d\u00e9ploiement se font dans le m\u00eame logiciel : pas d'outils \u00e0 assembler, pas de donn\u00e9es \u00e0 faire transiter.<\/p>\n    <\/div>\n    <div class=\"steps reveal\">\n      <div class=\"step\">\n        <div class=\"sn\">01<\/div>\n        <div class=\"mod\">Live \u00b7 Browser<\/div>\n        <h3>Acqu\u00e9rir les images<\/h3>\n        <p>En direct depuis les cam\u00e9ras de la ligne, ou \u00e0 distance depuis un t\u00e9l\u00e9phone ou une tablette pour les premiers essais. Les images sont rang\u00e9es en collections.<\/p>\n      <\/div>\n      <div class=\"step\">\n        <div class=\"sn\">02<\/div>\n        <div class=\"mod\">Dataset<\/div>\n        <h3>Annoter<\/h3>\n        <p>Les d\u00e9fauts et les pi\u00e8ces de r\u00e9f\u00e9rence sont annot\u00e9s directement dans l'outil, avec augmentation des donn\u00e9es pour enrichir les jeux d'images.<\/p>\n      <\/div>\n      <div class=\"step\">\n        <div class=\"sn\">03<\/div>\n        <div class=\"mod\">Training \u00b7 Models<\/div>\n        <h3>Entra\u00eener<\/h3>\n        <p>Entra\u00eenement des mod\u00e8les en ligne, suivi des essais et comparaison des versions : on promeut le meilleur mod\u00e8le en production.<\/p>\n      <\/div>\n      <div class=\"step\">\n        <div class=\"sn\">04<\/div>\n        <div class=\"mod\">Pipeline<\/div>\n        <h3>Assembler le contr\u00f4le<\/h3>\n        <p>Les briques sont reli\u00e9es dans l'\u00e9diteur visuel : pr\u00e9traitement, mod\u00e8les, mesures, r\u00e8gles de d\u00e9cision.<\/p>\n      <\/div>\n      <div class=\"step\">\n        <div class=\"sn\">05<\/div>\n        <div class=\"mod\">Simulation<\/div>\n        <h3>Valider sur images r\u00e9elles<\/h3>\n        <p>Le pipeline est rejou\u00e9 sur les collections enregistr\u00e9es pour r\u00e9gler les seuils et mesurer les performances avant la mise en ligne.<\/p>\n      <\/div>\n      <div class=\"step\">\n        <div class=\"sn\">06<\/div>\n        <div class=\"mod\">Layouts \u00b7 Deploy<\/div>\n        <h3>D\u00e9ployer<\/h3>\n        <p>On compose l'\u00e9cran op\u00e9rateur (flux vid\u00e9o, r\u00e9sultats, r\u00e9glages autoris\u00e9s) et on d\u00e9ploie sur le poste de la ligne, reli\u00e9 \u00e0 l'automate.<\/p>\n      <\/div>\n    <\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== VLM + REFLETS ===== -->\n<section class=\"sec\" id=\"innovations\" style=\"padding-top:0\">\n  <div class=\"wrap\">\n    <div class=\"sec-head reveal\">\n      <span class=\"eyebrow red\">Ce qui change la donne<\/span>\n      <h2>Des contr\u00f4les qui \u00e9taient hors de port\u00e9e.<\/h2>\n    <\/div>\n    <div class=\"duo\">\n      <article class=\"spot reveal\">\n        <div class=\"img\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27340%27%20viewBox%3D%270%200%20560%20340%27%3E%3Crect%20width%3D%27560%27%20height%3D%27340%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/cas-croissants.jpg\" alt=\"Contr\u00f4le de conformit\u00e9 de croissants par mod\u00e8le vision-langage dans Deeper\" loading=\"lazy\" title=\"\"><span class=\"lbl\">VLM \u00b7 vision par prompt<\/span><\/div>\n        <div class=\"body\">\n          <h3>D\u00e9marrer un contr\u00f4le sans campagne d'entra\u00eenement.<\/h3>\n          <p>Les mod\u00e8les vision-langage comprennent l'image et le texte. Sur la conformit\u00e9 de croissants, deux libell\u00e9s \u2014 \u00ab bien enroul\u00e9 \u00bb et \u00ab mal enroul\u00e9 \u00bb \u2014 ont suffi \u00e0 trier les produits, sans cr\u00e9er de mod\u00e8le ni annoter d'images.<\/p>\n          <p>Le VLM n'est pas une baguette magique : c'est un outil de plus dans la bo\u00eete. Selon le contr\u00f4le, Deeper le combine avec des mod\u00e8les entra\u00een\u00e9s et des outils de mesure pour atteindre la fiabilit\u00e9 attendue en production.<\/p>\n          <div class=\"chips\"><span>Mise en service rapide<\/span><span>OCR intelligent<\/span><span>Question \/ r\u00e9ponse<\/span><\/div>\n        <\/div>\n      <\/article>\n      <article class=\"spot reveal\">\n        <div class=\"img\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%271000%27%20height%3D%27511%27%20viewBox%3D%270%200%201000%20511%27%3E%3Crect%20width%3D%271000%27%20height%3D%27511%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/deeper-polar-w.jpg\" alt=\"Image en polarisation dans Deeper r\u00e9v\u00e9lant un d\u00e9faut sur une surface brillante\" loading=\"lazy\" title=\"\"><span class=\"lbl\">Polarisation \u00b7 retraitement<\/span><\/div>\n        <div class=\"body\">\n          <h3>Les reflets ne sont plus un obstacle.<\/h3>\n          <p>Verre, laque, films plastiques, m\u00e9tal poli : les surfaces brillantes mettent en \u00e9chec la plupart des syst\u00e8mes de vision. Avec une cam\u00e9ra polaris\u00e9e, Deeper calcule les param\u00e8tres de Stokes, supprime les reflets et fait ressortir bosses, rayures et poussi\u00e8res.<\/p>\n          <p>Sur des stylos laqu\u00e9s, des d\u00e9fauts inf\u00e9rieurs \u00e0 0,1 mm deviennent visibles ; sur des barquettes film\u00e9es, l'OCR relit dates et lots imprim\u00e9s \u00e0 travers le film.<\/p>\n          <div class=\"chips\"><span>Anti-reflet<\/span><span>Micro-d\u00e9fauts<\/span><span>OCR sous film<\/span><\/div>\n        <\/div>\n      <\/article>\n    <\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== SPECS ===== -->\n<section class=\"sec specs-sec\" id=\"specs\">\n  <div class=\"wrap\">\n    <div class=\"sec-head reveal\">\n      <span class=\"eyebrow\">Caract\u00e9ristiques<\/span>\n      <h2>Fiche technique.<\/h2>\n    <\/div>\n    <div class=\"specs reveal\">\n      <div class=\"spec\"><div class=\"k\">Calcul<\/div><div class=\"v\">PC industriel avec GPU<small>Forte puissance de calcul pour suivre les cadences de production<\/small><\/div><\/div>\n      <div class=\"spec\"><div class=\"k\">D\u00e9ploiement<\/div><div class=\"v\">En local sur PC ou en SaaS (serveur web)<small>En local, les images restent sur votre site<\/small><\/div><\/div>\n      <div class=\"spec\"><div class=\"k\">Cam\u00e9ras<\/div><div class=\"v\">Baumer, Basler, SVS-Vistek\u2026<small>Cam\u00e9ras matricielles, polaris\u00e9es ; r\u00e9glage exposition, gain, balance des blancs par preset<\/small><\/div><\/div>\n      <div class=\"spec\"><div class=\"k\">Multicam\u00e9ra<\/div><div class=\"v\">Plusieurs cam\u00e9ras g\u00e9r\u00e9es par une m\u00eame installation<small>Jusqu'\u00e0 8 cam\u00e9ras en flux sur nos applications verrerie<\/small><\/div><\/div>\n      <div class=\"spec\"><div class=\"k\">Programmation<\/div><div class=\"v\">\u00c9diteur de pipelines no-code<small>Plus de 80 briques, groupes, presets r\u00e9utilisables<\/small><\/div><\/div>\n      <div class=\"spec\"><div class=\"k\">Apprentissage<\/div><div class=\"v\">Annotation, entra\u00eenement et versionnage des mod\u00e8les en ligne<small>Promotion du meilleur mod\u00e8le en production<\/small><\/div><\/div>\n      <div class=\"spec\"><div class=\"k\">Interface op\u00e9rateur<\/div><div class=\"v\">\u00c9crans composables par glisser-d\u00e9poser<small>Flux vid\u00e9o, tableau de r\u00e9sultats, s\u00e9lection de r\u00e9f\u00e9rence, r\u00e9glages autoris\u00e9s<\/small><\/div><\/div>\n      <div class=\"spec\"><div class=\"k\">Communication<\/div><div class=\"v\">OPC-UA, Modbus TCP, requ\u00eates HTTP<small>Verdict OK\/NOK, mesures et textes lus transmis \u00e0 l'automate ou au SI<\/small><\/div><\/div>\n      <div class=\"spec\"><div class=\"k\">Mise au point<\/div><div class=\"v\">Mode simulation sur images enregistr\u00e9es<small>Visualisation de chaque \u00e9tape interm\u00e9diaire du pipeline<\/small><\/div><\/div>\n      <div class=\"spec\"><div class=\"k\">Acquisition \u00e0 distance<\/div><div class=\"v\">R\u00e9ception d'images depuis t\u00e9l\u00e9phone ou tablette<small>Pratique pour les \u00e9tudes de faisabilit\u00e9<\/small><\/div><\/div>\n    <\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== APPLICATIONS ===== -->\n<section class=\"sec\" id=\"applications\">\n  <div class=\"wrap\">\n    <div class=\"sec-head reveal\">\n      <span class=\"eyebrow red\">Deeper en production<\/span>\n      <h2>Des probl\u00e9matiques tr\u00e8s vari\u00e9es.<\/h2>\n      <p>Verrerie, cosm\u00e9tique, bois, agroalimentaire, emballage : quelques applications en service chez nos clients.<\/p>\n    <\/div>\n    <div class=\"cases\">\n      <article class=\"case reveal\">\n        <div class=\"ci\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27340%27%20viewBox%3D%270%200%20560%20340%27%3E%3Crect%20width%3D%27560%27%20height%3D%27340%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/cas-flacons.jpg\" alt=\"Inspection des fonds de flacons en verre\" loading=\"lazy\" title=\"\"><\/div>\n        <div class=\"cb\"><div class=\"sector\">Verrerie \u00b7 parfum & spiritueux<\/div><h3>Inspection de fonds de flacons<\/h3><p>Incis\u00e9s, bouillons et calcin\u00e9s d\u00e9tect\u00e9s sur toute la p\u00e9riph\u00e9rie du fond, en distinguant les vrais d\u00e9fauts des joints de moule.<\/p><div class=\"kpis\"><span>8 cam\u00e9ras<\/span><span>120 flacons\/min<\/span><span>~100 r\u00e9f.<\/span><\/div><\/div>\n      <\/article>\n      <article class=\"case reveal\">\n        <div class=\"ci\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27340%27%20viewBox%3D%270%200%20560%20340%27%3E%3Crect%20width%3D%27560%27%20height%3D%27340%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/cas-stylos.jpg\" alt=\"D\u00e9tection de micro-d\u00e9fauts sur stylos laqu\u00e9s\" loading=\"lazy\" title=\"\"><\/div>\n        <div class=\"cb\"><div class=\"sector\">Cosm\u00e9tique \u00b7 laquage<\/div><h3>Micro-d\u00e9fauts sur stylos laqu\u00e9s<\/h3><p>Le retraitement d'image supprime les reflets et met en relief bosses, poussi\u00e8res, rayures et d\u00e9fauts de laque.<\/p><div class=\"kpis\"><span>< 0,1 mm<\/span><span>Retraitement d'image<\/span><\/div><\/div>\n      <\/article>\n      <article class=\"case reveal\">\n        <div class=\"ci\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27340%27%20viewBox%3D%270%200%20560%20340%27%3E%3Crect%20width%3D%27560%27%20height%3D%27340%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/cas-bois.jpg\" alt=\"D\u00e9tection de n\u0153uds sur planches de bois\" loading=\"lazy\" title=\"\"><\/div>\n        <div class=\"cb\"><div class=\"sector\">Bois \u00b7 sortie de raboteuse<\/div><h3>Contr\u00f4le de planches<\/h3><p>N\u0153uds, fissures et manques de mati\u00e8re qualifi\u00e9s en temps r\u00e9el, jusqu'\u00e0 distinguer les n\u0153uds \u00ab moustaches \u00bb des n\u0153uds classiques.<\/p><div class=\"kpis\"><span>95 % de d\u00e9fauts d\u00e9tect\u00e9s<\/span><span>< 3 % faux positifs<\/span><\/div><\/div>\n      <\/article>\n      <article class=\"case reveal\">\n        <div class=\"ci\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27340%27%20viewBox%3D%270%200%20560%20340%27%3E%3Crect%20width%3D%27560%27%20height%3D%27340%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/cas-fromages.jpg\" alt=\"Contr\u00f4le multi-d\u00e9fauts de fromages\" loading=\"lazy\" title=\"\"><\/div>\n        <div class=\"cb\"><div class=\"sector\">Agroalimentaire \u00b7 fromage<\/div><h3>Fromages nus<\/h3><p>Corps \u00e9trangers, fissures, manques de mati\u00e8re, d\u00e9fauts de dimension et de fleur, sur produits en d\u00e9filement.<\/p><div class=\"kpis\"><span>80 fromages\/min<\/span><span>Multi-d\u00e9fauts<\/span><\/div><\/div>\n      <\/article>\n      <article class=\"case reveal\">\n        <div class=\"ci\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27340%27%20viewBox%3D%270%200%20560%20340%27%3E%3Crect%20width%3D%27560%27%20height%3D%27340%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/cas-poulet.jpg\" alt=\"D\u00e9tection d&#039;h\u00e9morragies sur manchons de poulet\" loading=\"lazy\" title=\"\"><\/div>\n        <div class=\"cb\"><div class=\"sector\">Volaille \u00b7 manchons<\/div><h3>Manchons & ailes de poulet<\/h3><p>H\u00e9morragies, h\u00e9matomes, os apparents, d\u00e9fauts de forme : chaque pi\u00e8ce class\u00e9e sans ralentir la ligne.<\/p><div class=\"kpis\"><span>~5 000 pi\u00e8ces\/h<\/span><span>93 % de d\u00e9fauts d\u00e9tect\u00e9s<\/span><\/div><\/div>\n      <\/article>\n      <article class=\"case reveal\">\n        <div class=\"ci\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27340%27%20viewBox%3D%270%200%20560%20340%27%3E%3Crect%20width%3D%27560%27%20height%3D%27340%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/cas-ocr-film.jpg\" alt=\"Lecture OCR de dates et lots sur barquettes film\u00e9es\" loading=\"lazy\" title=\"\"><\/div>\n        <div class=\"cb\"><div class=\"sector\">Charcuterie \u00b7 emballage<\/div><h3>OCR sur films plastiques<\/h3><p>Polarisation pour supprimer les reflets du film, puis IA capable de relire une impression jet d'encre irr\u00e9guli\u00e8re et les codes-barres.<\/p><div class=\"kpis\"><span>Polarisation<\/span><span>OCR + code-barres<\/span><\/div><\/div>\n      <\/article>\n      <article class=\"case reveal\">\n        <div class=\"ci\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27340%27%20viewBox%3D%270%200%20560%20340%27%3E%3Crect%20width%3D%27560%27%20height%3D%27340%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/cas-sandwich.jpg\" alt=\"Contr\u00f4le de compl\u00e9tude de sandwichs\" loading=\"lazy\" title=\"\"><\/div>\n        <div class=\"cb\"><div class=\"sector\">Agroalimentaire \u00b7 traiteur<\/div><h3>Sandwichs club<\/h3><p>V\u00e9rification que chaque sandwich est complet : aucune tranche ni garniture manquante.<\/p><div class=\"kpis\"><span>Manque de tranche<\/span><span>Manque de mati\u00e8re<\/span><\/div><\/div>\n      <\/article>\n      <article class=\"case reveal\">\n        <div class=\"ci\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%27560%27%20height%3D%27340%27%20viewBox%3D%270%200%20560%20340%27%3E%3Crect%20width%3D%27560%27%20height%3D%27340%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/cas-ocr.jpg\" alt=\"Lecture de caract\u00e8res grav\u00e9s sur une pi\u00e8ce cylindrique\" loading=\"lazy\" title=\"\"><\/div>\n        <div class=\"cb\"><div class=\"sector\">Tra\u00e7abilit\u00e9 \u00b7 marquage<\/div><h3>Lecture de marquages difficiles<\/h3><p>Caract\u00e8res grav\u00e9s ou imprim\u00e9s sur surfaces courbes et brillantes, relus caract\u00e8re par caract\u00e8re.<\/p><div class=\"kpis\"><span>OCR<\/span><span>Surfaces courbes<\/span><\/div><\/div>\n      <\/article>\n      <article class=\"case reveal\">\n        <div class=\"ci\"><img class=\"lazyload\" decoding=\"async\" src=\"data:image\/svg+xml,%3Csvg%20xmlns%3D%27http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%27%20width%3D%271000%27%20height%3D%27941%27%20viewBox%3D%270%200%201000%20941%27%3E%3Crect%20width%3D%271000%27%20height%3D%27941%27%20fill-opacity%3D%220%22%2F%3E%3C%2Fsvg%3E\" data-orig-src=\"https:\/\/www.cim-atlantique.com\/wp-content\/uploads\/2026\/10\/deeper-petri-w.jpg\" alt=\"Comptage automatique de colonies dans une bo\u00eete de Petri\" loading=\"lazy\" title=\"\"><\/div>\n        <div class=\"cb\"><div class=\"sector\">Laboratoire \u00b7 comptage<\/div><h3>Comptage automatique<\/h3><p>D\u00e9nombrement d'\u00e9l\u00e9ments sur une image, sans r\u00e8gle \u00e0 programmer, \u00e0 partir de quelques exemples.<\/p><div class=\"kpis\"><span>Comptage few-shot<\/span><span>Localisation<\/span><\/div><\/div>\n      <\/article>\n    <\/div>\n    <div class=\"cases-more reveal\">\n      <span>Plus de d\u00e9tails sur nos projets sur la page <a href=\"https:\/\/www.cim-atlantique.com\/fr\/vision-industrielle\/\" class=\"teal\" style=\"font-weight:600\">vision industrielle<\/a>.<\/span>\n      <a href=\"#contact\" class=\"btn btn-ghost btn-arrow\">Soumettre mon cas<\/a>\n    <\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== OFFRES ===== -->\n<section class=\"sec offres-sec\" id=\"offres\">\n  <canvas id=\"neural2\" aria-hidden=\"true\"><\/canvas>\n  <div class=\"wrap\">\n    <div class=\"sec-head reveal\">\n      <span class=\"eyebrow light\">Un logiciel, un int\u00e9grateur<\/span>\n      <h2>Cl\u00e9 en main, ou entre vos mains.<\/h2>\n      <p>CIM Atlantique con\u00e7oit des machines sp\u00e9ciales, des cellules robotis\u00e9es et des syst\u00e8mes de vision depuis plus de 30 ans. Deeper s'utilise de deux fa\u00e7ons.<\/p>\n    <\/div>\n    <div class=\"offres\">\n      <div class=\"offre main reveal\">\n        <span class=\"ot\">Solution cl\u00e9 en main<\/span>\n        <h3>Nous concevons et int\u00e9grons votre contr\u00f4le.<\/h3>\n        <p>Nos ing\u00e9nieurs ma\u00eetrisent toute la cha\u00eene, de l'\u00e9tude de faisabilit\u00e9 \u00e0 la mise en service sur site.<\/p>\n        <ul>\n          <li>\u00c9tude de faisabilit\u00e9 sur vos images et vos pi\u00e8ces<\/li>\n          <li>Choix de l'\u00e9clairage, de l'optique et de la cam\u00e9ra<\/li>\n          <li>D\u00e9veloppement de l'application Deeper<\/li>\n          <li>Int\u00e9gration m\u00e9canique, automatisme, \u00e9jection, robotique<\/li>\n          <li>Mise en service et suivi des performances<\/li>\n        <\/ul>\n        <a href=\"#contact\" class=\"btn btn-primary btn-arrow\">Demander une \u00e9tude<\/a>\n      <\/div>\n      <div class=\"offre reveal\">\n        <span class=\"ot\">Licence & formation<\/span>\n        <h3>Vos \u00e9quipes prennent la main.<\/h3>\n        <p>Pour les int\u00e9grateurs, les services m\u00e9thodes et les \u00e9quipes qualit\u00e9 qui veulent cr\u00e9er et faire \u00e9voluer leurs contr\u00f4les en autonomie.<\/p>\n        <ul>\n          <li>Licence du logiciel Deeper<\/li>\n          <li>Formation de prise en main \u00e0 distance, sur 2 \u00e0 3 jours<\/li>\n          <li>Formation finan\u00e7able par votre OPCO, via notre organisme partenaire certifi\u00e9 Qualiopi<\/li>\n          <li>Accompagnement de nos ing\u00e9nieurs sur vos premi\u00e8res applications<\/li>\n        <\/ul>\n        <a href=\"#contact\" class=\"btn btn-ghost on-dark btn-arrow\">\u00catre recontact\u00e9<\/a>\n      <\/div>\n    <\/div>\n    <div class=\"cir reveal\"><span class=\"pill\">Agr\u00e9\u00e9 CIR<\/span><span><b>Cr\u00e9dit d'Imp\u00f4t Recherche :<\/b> vos projets de vision innovants peuvent \u00eatre valoris\u00e9s.<\/span><\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== FAQ ===== -->\n<section class=\"sec\" id=\"faq\">\n  <div class=\"wrap faq-grid\">\n    <div class=\"sec-head reveal\">\n      <span class=\"eyebrow\">FAQ<\/span>\n      <h2>Questions fr\u00e9quentes sur Deeper.<\/h2>\n      <p>Une question qui n'est pas ici ? Appelez-nous au <a href=\"tel:+33240587296\" class=\"teal\" style=\"font-weight:600\">02 40 58 72 96<\/a>.<\/p>\n    <\/div>\n    <div class=\"faq reveal\">\n      <details open>\n        <summary>Deeper remplace-t-il Deep Capture ?<\/summary>\n        <p>Oui. Deeper est la nouvelle g\u00e9n\u00e9ration de notre logiciel de vision par IA. Il reprend l'exp\u00e9rience acquise avec Deep Capture en deep learning supervis\u00e9 et l'\u00e9largit : d\u00e9tection d'anomalies, mesure, OCR, comptage, polarisation, mod\u00e8les vision-langage, le tout programmable sans code. Vous avez une installation Deep Capture ? Parlons de son \u00e9volution.<\/p>\n      <\/details>\n      <details>\n        <summary>Faut-il \u00eatre expert en vision ou en IA ?<\/summary>\n        <p>Non. Les contr\u00f4les se construisent sans code, en assemblant des briques dans un \u00e9diteur visuel. Nos ing\u00e9nieurs peuvent livrer l'application cl\u00e9 en main, ou former vos \u00e9quipes pour qu'elles la fassent \u00e9voluer elles-m\u00eames.<\/p>\n      <\/details>\n      <details>\n        <summary>Combien d'images faut-il pour d\u00e9marrer ?<\/summary>\n        <p>Cela d\u00e9pend de la technique retenue. La d\u00e9tection d'anomalies apprend surtout \u00e0 partir de pi\u00e8ces bonnes, ce qui est pr\u00e9cieux quand les d\u00e9fauts sont rares. Les mod\u00e8les vision-langage permettent parfois de d\u00e9marrer sans entra\u00eenement sp\u00e9cifique. Pour les mod\u00e8les supervis\u00e9s, nous d\u00e9finissons avec vous la campagne d'images lors de l'\u00e9tude de faisabilit\u00e9.<\/p>\n      <\/details>\n      <details>\n        <summary>Quelles cam\u00e9ras sont compatibles ?<\/summary>\n        <p>Deeper fonctionne notamment avec les cam\u00e9ras industrielles Baumer, Basler et SVS-Vistek, y compris en polarisation. Il peut aussi recevoir des images \u00e0 distance (t\u00e9l\u00e9phone, tablette) pour la phase de mise au point.<\/p>\n      <\/details>\n      <details>\n        <summary>Cloud ou local ?<\/summary>\n        <p>Les deux : en local sur un PC \u00e9quip\u00e9 d'un GPU, au plus pr\u00e8s de la ligne et sans d\u00e9pendre d'internet, ou en SaaS sur un serveur web.<\/p>\n      <\/details>\n      <details>\n        <summary>Comment Deeper communique-t-il avec mon automate ?<\/summary>\n        <p>Le r\u00e9sultat du contr\u00f4le (verdict OK\/NOK, mesures, texte lu) est transmis \u00e0 l'automate ou au syst\u00e8me d'information via OPC-UA, Modbus TCP ou requ\u00eate HTTP.<\/p>\n      <\/details>\n      <details>\n        <summary>Peut-on tester Deeper sur nos pi\u00e8ces ?<\/summary>\n        <p>Oui. Envoyez-nous quelques images de vos pi\u00e8ces, bonnes et d\u00e9fectueuses : nous \u00e9valuons la faisabilit\u00e9 avec Deeper et vous pr\u00e9sentons les r\u00e9sultats, en visio ou dans nos ateliers de Grandchamps-des-Fontaines.<\/p>\n      <\/details>\n    <\/div>\n  <\/div>\n<\/section>\n\n<!-- ===== CONTACT ===== -->\n<section id=\"contact\">\n  <div class=\"wrap\">\n    <div class=\"contact reveal\">\n      <canvas id=\"neural3\" aria-hidden=\"true\"><\/canvas>\n      <div>\n        <span class=\"eyebrow light\">Et votre ligne ?<\/span>\n        <h2>Un contr\u00f4le difficile ? Testons-le avec Deeper.<\/h2>\n        <p class=\"lead\">Reflets, d\u00e9fauts al\u00e9atoires, grande vari\u00e9t\u00e9 de r\u00e9f\u00e9rences : envoyez-nous quelques images, nous vous disons ce que Deeper peut faire.<\/p>\n        <div class=\"cinfo\">\n          <div class=\"ci2\"><span class=\"k\">T\u00e9l\u00e9phone<\/span><a class=\"v\" href=\"tel:+33240587296\">02 40 58 72 96<\/a><\/div>\n          <div class=\"ci2\"><span class=\"k\">E-mail<\/span><a class=\"v\" href=\"mailto:r.gueudre@cim-atlantique.com\">r.gueudre@cim-atlantique.com<\/a><\/div>\n          <div class=\"ci2\"><span class=\"k\">Atelier<\/span><span class=\"v\">3 rue Herm\u00e8s, ZAC de la Belle \u00c9toile<br>44119 Grandchamps-des-Fontaines<\/span><\/div>\n        <\/div>\n      <\/div>\n      <form class=\"form\" id=\"demoForm\">\n        <h3>Demander une d\u00e9monstration<\/h3>\n        <p class=\"fh\">Sur vos propres pi\u00e8ces, en visio ou dans nos ateliers.<\/p>\n        <div class=\"field\"><label for=\"f-nom\">Nom et soci\u00e9t\u00e9<\/label><input id=\"f-nom\" name=\"nom\" required autocomplete=\"organization\"><\/div>\n        <div class=\"field\"><label for=\"f-mail\">E-mail professionnel<\/label><input id=\"f-mail\" name=\"mail\" type=\"email\" required autocomplete=\"email\"><\/div>\n        <div class=\"field\"><label for=\"f-tel\">T\u00e9l\u00e9phone<\/label><input id=\"f-tel\" name=\"tel\" type=\"tel\" autocomplete=\"tel\"><\/div>\n        <div class=\"field\"><label for=\"f-msg\">Votre contr\u00f4le<\/label><textarea id=\"f-msg\" name=\"msg\" placeholder=\"Produit, d\u00e9fauts \u00e0 d\u00e9tecter, cadence\u2026\"><\/textarea><\/div>\n        <button class=\"btn btn-primary btn-arrow\" type=\"submit\">Envoyer ma demande<\/button>\n        <p class=\"note\">Vos images pourront \u00eatre jointes \u00e0 notre r\u00e9ponse par e-mail.<\/p>\n      <\/form>\n    <\/div>\n  <\/div>\n<\/section>\n<\/div>\n<script>\n\/\/ reveal\n  if('IntersectionObserver' in window){\n    var io=new IntersectionObserver(function(es){\n      es.forEach(function(e){ if(e.isIntersecting){ e.target.classList.add('in'); io.unobserve(e.target); } });\n    },{threshold:.1, rootMargin:'0px 0px -40px 0px'});\n    document.querySelectorAll('.deeper-page .reveal').forEach(function(el){ io.observe(el); });\n  } else { document.querySelectorAll('.deeper-page .reveal').forEach(function(el){ el.classList.add('in'); }); }\n\n  \/\/ counters\n  function animateCount(el){\n    var target=parseFloat(el.getAttribute('data-count')), prefix=el.getAttribute('data-prefix')||'';\n    var unit=el.querySelector('.u'), unitHTML=unit?unit.outerHTML:'', dur=1300, start=null;\n    function step(ts){\n      if(!start)start=ts;\n      var p=Math.min((ts-start)\/dur,1), e=1-Math.pow(1-p,3);\n      el.innerHTML=prefix+Math.round(target*e)+unitHTML;\n      if(p<1)requestAnimationFrame(step);\n    }\n    requestAnimationFrame(step);\n  }\n  if('IntersectionObserver' in window){\n    var cio=new IntersectionObserver(function(es){\n      es.forEach(function(e){ if(e.isIntersecting){ animateCount(e.target); cio.unobserve(e.target); } });\n    },{threshold:.4});\n    document.querySelectorAll('.deeper-page [data-count]').forEach(function(el){ cio.observe(el); });\n  }\n\n  \/\/ neural network background : points libres comme avant, mais chaque point reste\n  \/\/ toujours reli\u00e9 \u00e0 ses 2 plus proches voisins -> aucune structure de moins de 3 points\n  function neural(canvas, opts){\n    if(!canvas) return;\n    var ctx=canvas.getContext('2d'), W,H, DPR=Math.min(window.devicePixelRatio||1,2);\n    var TEAL='29,145,158', RED='127,25,26', LIGHT=opts.light, LINE=LIGHT?'111,210,221':TEAL;\n    var nodes=[], N=opts.count||30, maxDist=opts.dist||170, MAXA=LIGHT?.35:.22, MINA=LIGHT?.12:.08;\n    var reduce=window.matchMedia&&window.matchMedia('(prefers-reduced-motion: reduce)').matches;\n    function resize(){ W=canvas.clientWidth; H=canvas.clientHeight; canvas.width=W*DPR; canvas.height=H*DPR; ctx.setTransform(DPR,0,0,DPR,0,0); }\n    function init(){ nodes=[]; for(var i=0;i<N;i++){ nodes.push({x:Math.random()*W,y:Math.random()*H,vx:(Math.random()-.5)*.35,vy:(Math.random()-.5)*.35,r:Math.random()*2+1.4,c:Math.random()<.18?RED:TEAL}); } }\n    function draw(){\n      ctx.clearRect(0,0,W,H); ctx.lineWidth=1;\n      var i,j,d,dx,dy,links={};\n      for(i=0;i<N;i++){ var a=nodes[i];\n        if(!reduce){ a.x+=a.vx; a.y+=a.vy; if(a.x<0||a.x>W)a.vx*=-1; if(a.y<0||a.y>H)a.vy*=-1; }\n      }\n      for(i=0;i<N;i++){\n        \/\/ liens par proximit\u00e9, comme avant\n        for(j=i+1;j<N;j++){ dx=nodes[i].x-nodes[j].x; dy=nodes[i].y-nodes[j].y; d=Math.sqrt(dx*dx+dy*dy);\n          if(d<maxDist) links[i+'-'+j]=Math.max(MINA,(1-d\/maxDist)*MAXA); }\n        \/\/ + lien obligatoire vers les 2 voisins les plus proches\n        var near=[];\n        for(j=0;j<N;j++){ if(j===i) continue; dx=nodes[i].x-nodes[j].x; dy=nodes[i].y-nodes[j].y; near.push([dx*dx+dy*dy,j]); }\n        near.sort(function(p,q){return p[0]-q[0];});\n        for(var k=0;k<2&&k<near.length;k++){ var o=near[k][1], key=i<o?i+'-'+o:o+'-'+i, dd=Math.sqrt(near[k][0]);\n          var al=Math.max(MINA,(1-Math.min(dd,maxDist*1.6)\/(maxDist*1.6))*MAXA);\n          if(!(key in links)||links[key]<al) links[key]=al; }\n      }\n      for(var key in links){ var ij=key.split('-'), A=nodes[+ij[0]], B=nodes[+ij[1]];\n        ctx.strokeStyle='rgba('+LINE+','+links[key]+')'; ctx.beginPath(); ctx.moveTo(A.x,A.y); ctx.lineTo(B.x,B.y); ctx.stroke(); }\n      for(var m=0;m<N;m++){ var n=nodes[m]; ctx.fillStyle='rgba('+(LIGHT&&n.c===TEAL?'111,210,221':n.c)+','+(LIGHT?.8:.5)+')'; ctx.beginPath(); ctx.arc(n.x,n.y,n.r,0,Math.PI*2); ctx.fill(); }\n      if(!reduce) requestAnimationFrame(draw);\n    }\n    resize(); init(); draw();\n    window.addEventListener('resize', function(){ resize(); init(); if(reduce) draw(); });\n  }\n  neural(document.getElementById('neural'),{count:window.innerWidth<700?18:34,dist:170});\n  neural(document.getElementById('neural2'),{count:26,dist:160,light:true});\n  neural(document.getElementById('neural3'),{count:22,dist:150,light:true});\n\n  \/\/ form -> mail\n  document.getElementById('demoForm').addEventListener('submit', function(e){\n    e.preventDefault();\n    var f=e.target, body='Nom \/ soci\u00e9t\u00e9 : '+f.nom.value+'\\nE-mail : '+f.mail.value+'\\nT\u00e9l\u00e9phone : '+f.tel.value+'\\n\\nContr\u00f4le :\\n'+f.msg.value;\n    window.location.href='mailto:r.gueudre@cim-atlantique.com?subject='+encodeURIComponent('Demande de d\u00e9mo Deeper')+'&body='+encodeURIComponent(body);\n  });\n<\/script><\/div><\/div><\/div><\/div>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":2,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"100-width.php","meta":{"inline_featured_image":false,"footnotes":""},"class_list":["post-7093","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.cim-atlantique.com\/fr\/wp-json\/wp\/v2\/pages\/7093","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.cim-atlantique.com\/fr\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.cim-atlantique.com\/fr\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.cim-atlantique.com\/fr\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.cim-atlantique.com\/fr\/wp-json\/wp\/v2\/comments?post=7093"}],"version-history":[{"count":3,"href":"https:\/\/www.cim-atlantique.com\/fr\/wp-json\/wp\/v2\/pages\/7093\/revisions"}],"predecessor-version":[{"id":7096,"href":"https:\/\/www.cim-atlantique.com\/fr\/wp-json\/wp\/v2\/pages\/7093\/revisions\/7096"}],"wp:attachment":[{"href":"https:\/\/www.cim-atlantique.com\/fr\/wp-json\/wp\/v2\/media?parent=7093"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}