Spaces:

yxma
/

react-force-recovery

Running

App Files Files Community

yxma commited on Aug 6

Commit

ac68efd

verified ·

1 Parent(s): af805c6

force recovery: methods, evaluation, debug log

Browse files

Files changed (23) hide show

.gitattributes +18 -0
assets/debug/cnc_01.png +3 -0
assets/debug/cnc_04.png +3 -0
assets/debug/cnc_08.png +3 -0
assets/debug/cnc_11.png +3 -0
assets/debug/cnc_14.png +3 -0
assets/debug/cnc_17.png +3 -0
assets/debug/feats_01.png +3 -0
assets/debug/feats_04.png +3 -0
assets/debug/feats_08.png +3 -0
assets/debug/feats_11.png +3 -0
assets/debug/feats_14.png +3 -0
assets/debug/feats_17.png +3 -0
assets/debug/glowtact_01.png +3 -0
assets/debug/glowtact_04.png +3 -0
assets/debug/glowtact_08.png +3 -0
assets/debug/glowtact_11.png +3 -0
assets/debug/glowtact_14.png +3 -0
assets/debug/glowtact_17.png +3 -0
debug_pipeline.html +106 -0
debug_pipeline_zh.html +93 -0
results.html +4 -0
results_zh.html +2 -0

.gitattributes CHANGED Viewed

@@ -75,3 +75,21 @@ assets/glowtact_comparison.png filter=lfs diff=lfs merge=lfs -text
 assets/results_cnc.png filter=lfs diff=lfs merge=lfs -text
 assets/results_feats.png filter=lfs diff=lfs merge=lfs -text
 assets/results_glowtact.png filter=lfs diff=lfs merge=lfs -text

 assets/results_cnc.png filter=lfs diff=lfs merge=lfs -text
 assets/results_feats.png filter=lfs diff=lfs merge=lfs -text
 assets/results_glowtact.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/cnc_01.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/cnc_04.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/cnc_08.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/cnc_11.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/cnc_14.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/cnc_17.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/feats_01.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/feats_04.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/feats_08.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/feats_11.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/feats_14.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/feats_17.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/glowtact_01.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/glowtact_04.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/glowtact_08.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/glowtact_11.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/glowtact_14.png filter=lfs diff=lfs merge=lfs -text
+assets/debug/glowtact_17.png filter=lfs diff=lfs merge=lfs -text

assets/debug/cnc_01.png ADDED Viewed

Git LFS Details

SHA256: 961a043191d65375c9032984fd8a20a357ece3cf7c9adbca961a09daf24dcabe
Pointer size: 131 Bytes
Size of remote file: 333 kB

assets/debug/cnc_04.png ADDED Viewed

Git LFS Details

SHA256: e829cfe7d004a42230f3bdccada271b67b860da261a6f86c6438e7e921d00c9d
Pointer size: 131 Bytes
Size of remote file: 333 kB

assets/debug/cnc_08.png ADDED Viewed

Git LFS Details

SHA256: 2a6d936df49e4e2c36b322d5f7dc2eb319684a58db47bda2aac8d48cffc22b8b
Pointer size: 131 Bytes
Size of remote file: 308 kB

assets/debug/cnc_11.png ADDED Viewed

Git LFS Details

SHA256: 1306c16b2b44fc339a60c396e45fa716fbc091f3aeacf283bb3cc565f24afad8
Pointer size: 131 Bytes
Size of remote file: 343 kB

assets/debug/cnc_14.png ADDED Viewed

Git LFS Details

SHA256: e5f3f8104ea1b279434e6cb4b61d0d6054f383b686a43ae826037135ab5acbd6
Pointer size: 131 Bytes
Size of remote file: 319 kB

assets/debug/cnc_17.png ADDED Viewed

Git LFS Details

SHA256: 393911cbe8b82a3dc6cddf9e2678b28f4b3045a4ac8af5027a1ad0e5f4babefc
Pointer size: 131 Bytes
Size of remote file: 334 kB

assets/debug/feats_01.png ADDED Viewed

Git LFS Details

SHA256: 8edecbaa96a596379f4230a84d1ece29297513f547f034e7b7d088c6fd03b1a7
Pointer size: 131 Bytes
Size of remote file: 486 kB

assets/debug/feats_04.png ADDED Viewed

Git LFS Details

SHA256: 57f64b4929468aa3a7dae870644aa4bd91417534a6f2fd14aa78f97f27d8ca1e
Pointer size: 131 Bytes
Size of remote file: 499 kB

assets/debug/feats_08.png ADDED Viewed

Git LFS Details

SHA256: de6448a20f305e13a1d3ec622596ecef468e98a3e31eeb6a2c70b66157df1ea1
Pointer size: 131 Bytes
Size of remote file: 460 kB

assets/debug/feats_11.png ADDED Viewed

Git LFS Details

SHA256: e1c6f35f5efcc2434602cd15c52a682adc8ecb725d0058af1d718f3a1c298e27
Pointer size: 131 Bytes
Size of remote file: 474 kB

assets/debug/feats_14.png ADDED Viewed

Git LFS Details

SHA256: 7b6e4ee033748d6198ed05512602659b60cee1adda3fe8cb912e7063c630bd0d
Pointer size: 131 Bytes
Size of remote file: 470 kB

assets/debug/feats_17.png ADDED Viewed

Git LFS Details

SHA256: 3102c67a188e1dce5f62ee6993be00cdbc7c92b5cb470b882afc1f944bc7fbed
Pointer size: 131 Bytes
Size of remote file: 564 kB

assets/debug/glowtact_01.png ADDED Viewed

Git LFS Details

SHA256: f1a12ad0ee4a4914ad02da6b5bff07a7df8698ccebd07df928e2f9a47a5769e8
Pointer size: 131 Bytes
Size of remote file: 256 kB

assets/debug/glowtact_04.png ADDED Viewed

Git LFS Details

SHA256: b62a864d8dc45bc0f44d810d7abd0ba3b1837447e2bbc8bc934c2d9a5c097806
Pointer size: 131 Bytes
Size of remote file: 311 kB

assets/debug/glowtact_08.png ADDED Viewed

Git LFS Details

SHA256: b1eb7237c49a2d88adb345f4a78b8359509986ff49fe05cf07c95f135ab9ef55
Pointer size: 131 Bytes
Size of remote file: 267 kB

assets/debug/glowtact_11.png ADDED Viewed

Git LFS Details

SHA256: 7b334b4465bfd8174daf86cbcfc53120b644eebef3f26e84165d8c9e1ccc7da4
Pointer size: 131 Bytes
Size of remote file: 295 kB

assets/debug/glowtact_14.png ADDED Viewed

Git LFS Details

SHA256: b9cb2fa30c299ce1c916383fabcfcb82f823328fe0641192127110daac6c9960
Pointer size: 131 Bytes
Size of remote file: 337 kB

assets/debug/glowtact_17.png ADDED Viewed

Git LFS Details

SHA256: 3cc7f9be3f7949d9023e0173f64f50e41b2fd546369248a696efca22d1c0365b
Pointer size: 131 Bytes
Size of remote file: 358 kB

debug_pipeline.html ADDED Viewed

	@@ -0,0 +1,106 @@

+<!doctype html><html lang="en"><head><meta charset="utf-8">
+<meta name="viewport" content="width=device-width,initial-scale=1">
+<title>Pipeline debug — raw image to force, step by step</title>
+<style>
+:root{--paper:#0d1526;--ink:#dce7f5;--dim:#7c8db0;--grid:#1a2540;
+--force:#ffb347;--target:#ff7847;--ok:#7be0a0;--card:#111b31;--line:#24345;}
+*{box-sizing:border-box}
+body{margin:0;background:var(--paper);color:var(--ink);
+font-family:'IBM Plex Sans','Noto Sans SC',sans-serif;line-height:1.65;font-size:16px;
+background-image:linear-gradient(var(--grid) 1px,transparent 1px),
+linear-gradient(90deg,var(--grid) 1px,transparent 1px);background-size:44px 44px}
+.wrap{max-width:880px;margin:0 auto;padding:0 24px 80px}
+h1,h2{font-family:'Fraunces','Noto Serif SC',serif}
+h1{font-size:2.1rem;font-weight:700;line-height:1.2;margin:0 0 10px}
+h2{font-size:1.25rem;font-weight:600;margin:46px 0 8px;color:var(--force)}
+.kicker{font-family:'IBM Plex Mono',monospace;font-size:.76rem;letter-spacing:.16em;
+text-transform:uppercase;color:var(--force);margin-bottom:14px}
+header{padding:60px 0 26px;border-bottom:1px dashed var(--line)}
+.sub{color:var(--dim);max-width:720px}
+.card{background:var(--card);border:1px solid var(--line);border-radius:10px;
+padding:18px 22px;margin:14px 0}
+.step{display:grid;grid-template-columns:34px 200px 1fr;gap:12px;
+padding:10px 0;border-bottom:1px solid var(--line);align-items:baseline}
+.step:last-child{border-bottom:none}
+.step .n{font-family:'Fraunces',serif;font-size:1.25rem;color:var(--target)}
+.step .t{font-family:'IBM Plex Mono',monospace;font-size:.85rem}
+.step .d{color:var(--dim);font-size:.9rem}
+table{border-collapse:collapse;width:100%;font-size:.88rem;margin:10px 0;
+font-variant-numeric:tabular-nums}
+th,td{padding:7px 10px;border-bottom:1px solid var(--line);text-align:right}
+th{color:var(--dim);font-family:'IBM Plex Mono',monospace;font-size:.72rem;
+text-transform:uppercase;letter-spacing:.08em}
+td:first-child,th:first-child{text-align:left}
+img{max-width:100%;border-radius:8px;border:1px solid var(--line);display:block;margin:12px auto}
+a{color:#4fd8e0}a:hover{color:var(--force)}
+.pill{display:inline-block;border:1px solid var(--line);border-radius:999px;
+padding:3px 12px;font-size:.75rem;font-family:'IBM Plex Mono',monospace;
+color:var(--dim);margin-right:8px;margin-top:12px;text-decoration:none}
+code{font-family:'IBM Plex Mono',monospace;background:#0a1120;padding:1px 5px;
+border-radius:4px;font-size:.85em}
+.ok{color:var(--ok)}.warn{color:#e0b34f}
+footer{margin-top:60px;padding-top:18px;border-top:1px dashed var(--line);
+color:var(--dim);font-size:.8rem}
+@media(max-width:640px){.step{grid-template-columns:28px 1fr;grid-auto-flow:dense}
+.step .d{grid-column:2}
+h1{font-size:1.45rem}
+table{display:block;overflow-x:auto;white-space:nowrap;-webkit-overflow-scrolling:touch}
+th,td{white-space:normal;min-width:110px}}
+ .srow img{width:100%;max-width:1700px;background:#fff;
+border-radius:6px;margin:4px 0}</style></head><body><div class="wrap">
+<div class="kicker">React force recovery · pipeline debug
+<span style="float:right"><a href="debug_pipeline_zh.html">中文</a></span></div>
+<h1>From raw image to force, step by step</h1>
+<p>Each row: raw &rarr; reference &rarr; dI=img&minus;ref &rarr; valid mask
+&rarr; |LUT gradient| &rarr; Poisson depth &rarr; features and predicted vs
+ground-truth force. Same pipeline, three datasets.</p>
+<h2>Why cnc_Mini looked broken — and what it actually was</h2>
+<p>Neither bad labels nor a broken reconstruction: the cnc press grid spans
+the full 20&times;16 mm pad while the cropped camera view sees only
+&asymp;13&times;10 mm, so two thirds of the presses are partially outside the
+image and their contact features are systematically underestimated. Restricted
+to presses fully in view, the same reconstruction reaches GlowTact-level
+accuracy. Label quality is fine (force CV 4.5% at fixed probe/position/depth).
+This retracts our earlier &ldquo;dataset ceiling&rdquo; argument: in-view
+&rho;=0.94 exceeds the commanded-depth proxy (0.78&ndash;0.88) that argument
+relied on.</p>
+<table><tr><th>scope (cnc, 3351 frames)</th><th>n</th>
+<th>per-probe &rho; (A&ndash;F)</th><th>median &rho;</th><th>MAE</th></tr>
+<tr><td>all press positions</td><td>3351</td><td>0.09&ndash;0.16</td>
+<td>0.113</td><td>0.74 N</td></tr>
+<tr><td>interior x&isin;[3.5,14.5]</td><td>1232</td><td>0.79&ndash;0.85</td>
+<td>0.835</td><td>0.39 N</td></tr>
+<tr><td><b>strictly in view x&isin;[5,13]</b></td><td>671</td>
+<td><b>0.94 0.94 0.93 0.91 0.95 0.95</b></td><td><b>0.941</b></td>
+<td><b>0.26 N</b></td></tr></table>
+<p class="note">Per-probe half/half fit, 7 seeds, LUT pipeline with
+z-supervised gain field. Dot-type control: FEATS presses are always centered,
+so the raw pipeline already gives &rho;=0.72 there, vs 0.46 raw on GlowTact —
+the gain field and scope, not markers, are the binding factors.</p>
+<h2>GlowTact — markerless, the LUT's home sensor (control)</h2><div class='srow'>
+<img src='assets/debug/glowtact_01.png' loading='lazy'>
+<img src='assets/debug/glowtact_04.png' loading='lazy'>
+<img src='assets/debug/glowtact_08.png' loading='lazy'>
+<img src='assets/debug/glowtact_11.png' loading='lazy'>
+<img src='assets/debug/glowtact_14.png' loading='lazy'>
+<img src='assets/debug/glowtact_17.png' loading='lazy'>
+</div>
+<h2>FoTa cnc_Mini — markerless, foreign sensor, press grid larger than the field of view</h2><div class='srow'>
+<img src='assets/debug/cnc_01.png' loading='lazy'>
+<img src='assets/debug/cnc_04.png' loading='lazy'>
+<img src='assets/debug/cnc_08.png' loading='lazy'>
+<img src='assets/debug/cnc_11.png' loading='lazy'>
+<img src='assets/debug/cnc_14.png' loading='lazy'>
+<img src='assets/debug/cnc_17.png' loading='lazy'>
+</div>
+<h2>FEATS — dot/marker type; difference imaging cancels static markers</h2><div class='srow'>
+<img src='assets/debug/feats_01.png' loading='lazy'>
+<img src='assets/debug/feats_04.png' loading='lazy'>
+<img src='assets/debug/feats_08.png' loading='lazy'>
+<img src='assets/debug/feats_11.png' loading='lazy'>
+<img src='assets/debug/feats_14.png' loading='lazy'>
+<img src='assets/debug/feats_17.png' loading='lazy'>
+</div>
+<footer>React force recovery · <a href="index.html">overview</a> ·
+<a href="results.html">results</a> · <a href="method.html">method design</a>
+</footer></div></body></html>

debug_pipeline_zh.html ADDED Viewed

	@@ -0,0 +1,93 @@

+<!doctype html><html lang="zh-CN"><head><meta charset="utf-8">
+<meta name="viewport" content="width=device-width,initial-scale=1">
+<title>管线调试 — 从原始图像到力，逐步展示</title>
+<style>
+:root{--paper:#0d1526;--ink:#dce7f5;--dim:#7c8db0;--grid:#1a2540;
+--force:#ffb347;--target:#ff7847;--ok:#7be0a0;--card:#111b31;--line:#24345;}
+*{box-sizing:border-box}
+body{margin:0;background:var(--paper);color:var(--ink);
+font-family:'IBM Plex Sans','Noto Sans SC',sans-serif;line-height:1.65;font-size:16px;
+background-image:linear-gradient(var(--grid) 1px,transparent 1px),
+linear-gradient(90deg,var(--grid) 1px,transparent 1px);background-size:44px 44px}
+.wrap{max-width:880px;margin:0 auto;padding:0 24px 80px}
+h1,h2{font-family:'Fraunces','Noto Serif SC',serif}
+h1{font-size:2.1rem;font-weight:700;line-height:1.2;margin:0 0 10px}
+h2{font-size:1.25rem;font-weight:600;margin:46px 0 8px;color:var(--force)}
+.kicker{font-family:'IBM Plex Mono',monospace;font-size:.76rem;letter-spacing:.16em;
+text-transform:uppercase;color:var(--force);margin-bottom:14px}
+header{padding:60px 0 26px;border-bottom:1px dashed var(--line)}
+.sub{color:var(--dim);max-width:720px}
+.card{background:var(--card);border:1px solid var(--line);border-radius:10px;
+padding:18px 22px;margin:14px 0}
+.step{display:grid;grid-template-columns:34px 200px 1fr;gap:12px;
+padding:10px 0;border-bottom:1px solid var(--line);align-items:baseline}
+.step:last-child{border-bottom:none}
+.step .n{font-family:'Fraunces',serif;font-size:1.25rem;color:var(--target)}
+.step .t{font-family:'IBM Plex Mono',monospace;font-size:.85rem}
+.step .d{color:var(--dim);font-size:.9rem}
+table{border-collapse:collapse;width:100%;font-size:.88rem;margin:10px 0;
+font-variant-numeric:tabular-nums}
+th,td{padding:7px 10px;border-bottom:1px solid var(--line);text-align:right}
+th{color:var(--dim);font-family:'IBM Plex Mono',monospace;font-size:.72rem;
+text-transform:uppercase;letter-spacing:.08em}
+td:first-child,th:first-child{text-align:left}
+img{max-width:100%;border-radius:8px;border:1px solid var(--line);display:block;margin:12px auto}
+a{color:#4fd8e0}a:hover{color:var(--force)}
+.pill{display:inline-block;border:1px solid var(--line);border-radius:999px;
+padding:3px 12px;font-size:.75rem;font-family:'IBM Plex Mono',monospace;
+color:var(--dim);margin-right:8px;margin-top:12px;text-decoration:none}
+code{font-family:'IBM Plex Mono',monospace;background:#0a1120;padding:1px 5px;
+border-radius:4px;font-size:.85em}
+.ok{color:var(--ok)}.warn{color:#e0b34f}
+footer{margin-top:60px;padding-top:18px;border-top:1px dashed var(--line);
+color:var(--dim);font-size:.8rem}
+@media(max-width:640px){.step{grid-template-columns:28px 1fr;grid-auto-flow:dense}
+.step .d{grid-column:2}
+h1{font-size:1.45rem}
+table{display:block;overflow-x:auto;white-space:nowrap;-webkit-overflow-scrolling:touch}
+th,td{white-space:normal;min-width:110px}}
+ .srow img{width:100%;max-width:1700px;background:#fff;
+border-radius:6px;margin:4px 0}</style></head><body><div class="wrap">
+<div class="kicker">React force recovery · 管线调试
+<span style="float:right"><a href="debug_pipeline.html">English</a></span></div>
+<h1>从原始图像到力，逐步展示</h1>
+<p>每行：原图 &rarr; 参考帧 &rarr; 差分 dI=img&minus;ref &rarr; 有效掩码 &rarr; |LUT 梯度| &rarr; Poisson 深度 &rarr; 特征与预测力 vs 真值。同一管线，三个数据集。</p>
+<h2>cnc_Mini 为何看起来失效 — 真正的原因</h2>
+<p>既不是标签质量差，也不是重建失效：cnc 的按压网格覆盖整块 20&times;16 mm 凝胶，而裁剪后的相机视野只有约 13&times;10 mm——三分之二的按压部分在画面外，接触特征被系统性低估。仅取完全在视野内的按压，同一重建管线达到 GlowTact 同级精度。标签质量合格（固定探头/位置/深度下力的变异系数 4.5%）。这也撤回了我们此前的“数据集天花板”论证：视野内 &rho;=0.94 超过了该论证依赖的指令深度代理值（0.78&ndash;0.88）。</p>
+<table><tr><th>作用域（cnc，3351 帧）</th><th>n</th>
+<th>per-probe &rho; (A&ndash;F)</th><th>median &rho;</th><th>MAE</th></tr>
+<tr><td>全部按压位置</td><td>3351</td><td>0.09&ndash;0.16</td>
+<td>0.113</td><td>0.74 N</td></tr>
+<tr><td>内部 x&isin;[3.5,14.5]</td><td>1232</td><td>0.79&ndash;0.85</td>
+<td>0.835</td><td>0.39 N</td></tr>
+<tr><td><b>严格视野内 x&isin;[5,13]</b></td><td>671</td>
+<td><b>0.94 0.94 0.93 0.91 0.95 0.95</b></td><td><b>0.941</b></td>
+<td><b>0.26 N</b></td></tr></table>
+<p class="note">每探头半半分拟合，7 种子，LUT 管线 + z 监督增益场。dot 类型对照：FEATS 按压恒居中，原始管线即得 &rho;=0.72，而 GlowTact 原始条件仅 0.46——起决定作用的是增益场与作用域，而非 marker。</p>
+<h2>GlowTact — 无 marker，LUT 标定所用传感器（对照）</h2><div class='srow'>
+<img src='assets/debug/glowtact_01.png' loading='lazy'>
+<img src='assets/debug/glowtact_04.png' loading='lazy'>
+<img src='assets/debug/glowtact_08.png' loading='lazy'>
+<img src='assets/debug/glowtact_11.png' loading='lazy'>
+<img src='assets/debug/glowtact_14.png' loading='lazy'>
+<img src='assets/debug/glowtact_17.png' loading='lazy'>
+</div>
+<h2>FoTa cnc_Mini — 无 marker，外来传感器，按压网格大于视野</h2><div class='srow'>
+<img src='assets/debug/cnc_01.png' loading='lazy'>
+<img src='assets/debug/cnc_04.png' loading='lazy'>
+<img src='assets/debug/cnc_08.png' loading='lazy'>
+<img src='assets/debug/cnc_11.png' loading='lazy'>
+<img src='assets/debug/cnc_14.png' loading='lazy'>
+<img src='assets/debug/cnc_17.png' loading='lazy'>
+</div>
+<h2>FEATS — dot/marker 类型；差分成像抵消静态 marker</h2><div class='srow'>
+<img src='assets/debug/feats_01.png' loading='lazy'>
+<img src='assets/debug/feats_04.png' loading='lazy'>
+<img src='assets/debug/feats_08.png' loading='lazy'>
+<img src='assets/debug/feats_11.png' loading='lazy'>
+<img src='assets/debug/feats_14.png' loading='lazy'>
+<img src='assets/debug/feats_17.png' loading='lazy'>
+</div>
+<footer>React force recovery · <a href="index.html">overview</a> ·
+<a href="results.html">results</a> · <a href="method.html">method design</a>
+</footer></div></body></html>

results.html CHANGED Viewed

@@ -119,6 +119,10 @@ where no trained model matches the domain: the physics pipeline is the only
 option that works out of the box — and its FEATS-dataset score (0.74) shows
 what it does on a domain nobody tuned it for.</p></div>
 <footer>React force recovery · <a href="index.html">overview</a> ·
 <a href="method.html">method design</a> · <a href="gallery.html">gallery</a> ·
 data: FEATS (2411.03315) · FoTa/T3 (2406.13640) · GlowTact

 option that works out of the box — and its FEATS-dataset score (0.74) shows
 what it does on a domain nobody tuned it for.</p></div>
+<p><a href="debug_pipeline.html"><b>Pipeline debug page</b></a>: raw
+image &rarr; force step by step on all three datasets, and the cnc
+field-of-view ablation (in-view &rho;=0.94).</p>
 <footer>React force recovery · <a href="index.html">overview</a> ·
 <a href="method.html">method design</a> · <a href="gallery.html">gallery</a> ·
 data: FEATS (2411.03315) · FoTa/T3 (2406.13640) · GlowTact

results_zh.html CHANGED Viewed

@@ -110,6 +110,8 @@ FeelAnyForce(20 万帧)——一致性 ρ=0.91,且物理管线判无接触的每
 物理管线独立审计,分歧行打标。<b>换任何新 gel 或新传感器</b>、没有训练模型匹配该域时:
 物理管线是唯一开箱即用的选项——它在 FEATS 数据集上的 0.74 就是"无人为它调过的域"上的表现。</p></div>
 <footer>React 力恢复 · <a href="index.html">总览</a> ·
 <a href="method_zh.html">方法设计</a> · <a href="gallery.html">图库</a> ·
 数据:FEATS (2411.03315) · FoTa/T3 (2406.13640) · GlowTact

 物理管线独立审计,分歧行打标。<b>换任何新 gel 或新传感器</b>、没有训练模型匹配该域时:
 物理管线是唯一开箱即用的选项——它在 FEATS 数据集上的 0.74 就是"无人为它调过的域"上的表现。</p></div>
+<p><a href="debug_pipeline.html"><b>管线调试页</b></a>：三个数据集上从原始图像到力的逐步展示，以及 cnc 视野消融（视野内 &rho;=0.94）。</p>
 <footer>React 力恢复 · <a href="index.html">总览</a> ·
 <a href="method_zh.html">方法设计</a> · <a href="gallery.html">图库</a> ·
 数据:FEATS (2411.03315) · FoTa/T3 (2406.13640) · GlowTact