[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7BVVX9SDVNFJM3SBEPZKDZK":3,"skill-same:01M7BVVX9SDVNFJM3SBEPZKDZK":45},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M7BVVX9SDVNFJM3SBEPZKDZK","egomotion-estimation",null,"Estimate camera motion with optical flow + affine\u002Fhomography, allow multi-label per frame.","linyouxue\u002Fskill-repair-benchmark","linyouxue","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F216151464?v=4","01M7BJEVCXKD74KH6FG9CBVSQB","林宥学","01M7BJEVE02NP217V93BKZFC0J",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FMMG2Skill-Opus47-Gold15-20260924\u002Ftasks\u002Fdynamic-object-aware-egomotion\u002Frepaired_skill\u002Fegomotion-estimation\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: egomotion-estimation\ndescription: \"Estimate camera motion with optical flow + affine\u002Fhomography, allow multi-label per frame.\"\n---\n\n# When to use\n- You need to classify camera motion (Stay\u002FDolly\u002FPan\u002FTilt\u002FRoll) from video, allowing multiple labels on the same frame.\n# Workflow\n1) **Feature tracking**: `goodFeaturesToTrack` on `prev_gray` + `calcOpticalFlowPyrLK(prev→curr)`; drop tracks with status=0 or too few remaining.\n2) **Robust transform**: `estimateAffinePartial2D(prev_pts, curr_pts, method=RANSAC)` — this returns M mapping *prev→curr*, so `(dx,dy)=M[:,2]` is how features shifted in the image from prev to curr.\n3) **Sign convention (map to CAMERA motion, not image motion)**:\n   - Features moving RIGHT in image (`dx>0`) ⇒ camera panned LEFT ⇒ label `\"Pan Left\"`.\n   - Features moving LEFT (`dx\u003C0`) ⇒ `\"Pan Right\"`.\n   - Features moving DOWN (`dy>0`) ⇒ camera tilted UP ⇒ `\"Tilt Up\"`.\n   - Features moving UP (`dy\u003C0`) ⇒ `\"Tilt Down\"`.\n   - `scale>1` (features spread outward) ⇒ `\"Dolly In\"`; `scale\u003C1` ⇒ `\"Dolly Out\"`.\n   - `rotation>0` (CCW in image) ⇒ `\"Roll Right\"` for a hand-held camera rolling clockwise about its optical axis (verify on the specific transform direction you used; flip if needed).\n   Confirm the mapping by sanity-checking one frame with an obvious motion cue in the video (e.g. horizon shift) before finalizing.\n4) **Thresholding**: `th_trans` in px\u002Fframe (scale with resolution and fps step), `th_rot` in radians, `th_scale` as relative deviation from 1.\n5) **Multi-label**: independently test each axis; do NOT gate one axis on another (esp. do not require `|dx|>=|dy|` for Pan — that suppresses legitimate Tilt+Pan combos). Emit both if both exceed threshold.\n6) **Outlier \u002F transition preservation**: after per-frame labels, run mild temporal smoothing (window mode); do NOT collapse away isolated frames whose translation clearly exceeds threshold (e.g. `|dy|>th_trans`) — they should surface as a distinct interval.\n7) **Interval compression**: merge consecutive frames with identical label *sets* (order-insensitive) into `\"start->end\"` half-open ranges.\n8) **Fallback**: if the transform estimate fails, use identity → `\"Stay\"`; never emit an empty label list.\n# Decision sketch\n```python\nlabels=[]\nfor i in range(1, len(sample_ids)):\n    M, _ = cv2.estimateAffinePartial2D(prev_pts, curr_pts, method=cv2.RANSAC)\n    if M is None: labels.append([\"Stay\"]); continue\n    dx, dy = M[0,2], M[1,2]\n    a, b = M[0,0], M[1,0]\n    scale = math.hypot(a,b); rot = math.atan2(b,a)\n    lbl=[]\n    if abs(scale-1)>th_scale: lbl.append(\"Dolly In\" if scale>1 else \"Dolly Out\")\n    if abs(rot)>th_rot:       lbl.append(\"Roll Right\" if rot>0 else \"Roll Left\")\n    if abs(dx)>th_trans:      lbl.append(\"Pan Left\"  if dx>0 else \"Pan Right\")\n    if abs(dy)>th_trans:      lbl.append(\"Tilt Up\"   if dy>0 else \"Tilt Down\")\n    if not lbl: lbl.append(\"Stay\")\n    labels.append(lbl)\n```\n# Self-check\n- [ ] Fallback to identity transform on failure; never emit empty labels.\n- [ ] Sign convention explicitly matches CAMERA motion (image-right feature shift ⇒ Pan Left, etc.), and was sanity-checked against one obvious cue in the clip.\n- [ ] Pan and Tilt tested independently — no `|dx|>=|dy|` gating.\n- [ ] Frames whose translation crosses threshold are not silently merged away.\n- [ ] Compressed intervals cover all sampled frames as a half-open partition of `[0, n_samples)`.\n","# When to use\n- You need to classify camera motion (Stay\u002FDolly\u002FPan\u002FTilt\u002FRoll) from video, allowing multiple labels on the same frame.\n# Workflow\n1) **Feature tracking**: `goodFeaturesToTrack` on `prev_gray` + `calcOpticalFlowPyrLK(prev→curr)`; drop tracks with status=0 or too few remaining.\n2) **Robust transform**: `estimateAffinePartial2D(prev_pts, curr_pts, method=RANSAC)` — this returns M mapping *prev→curr*, so `(dx,dy)=M[:,2]` is how features shifted in the image from prev to curr.\n3) **Sign convention (map to CAMERA motion, not image motion)**:\n   - Features moving RIGHT in image (`dx>0`) ⇒ camera panned LEFT ⇒ label `\"Pan Left\"`.\n   - Features moving LEFT (`dx\u003C0`) ⇒ `\"Pan Right\"`.\n   - Features moving DOWN (`dy>0`) ⇒ camera tilted UP ⇒ `\"Tilt Up\"`.\n   - Features moving UP (`dy\u003C0`) ⇒ `\"Tilt Down\"`.\n   - `scale>1` (features spread outward) ⇒ `\"Dolly In\"`; `scale\u003C1` ⇒ `\"Dolly Out\"`.\n   - `rotation>0` (CCW in image) ⇒ `\"Roll Right\"` for a hand-held camera rolling clockwise about its optical axis (verify on the specific transform direction you used; flip if needed).\n   Confirm the mapping by sanity-checking one frame with an obvious motion cue in the video (e.g. horizon shift) before finalizing.\n4) **Thresholding**: `th_trans` in px\u002Fframe (scale with resolution and fps step), `th_rot` in radians, `th_scale` as relative deviation from 1.\n5) **Multi-label**: independently test each axis; do NOT gate one axis on another (esp. do not require `|dx|>=|dy|` for Pan — that suppresses legitimate Tilt+Pan combos). Emit both if both exceed threshold.\n6) **Outlier \u002F transition preservation**: after per-frame labels, run mild temporal smoothing (window mode); do NOT collapse away isolated frames whose translation clearly exceeds threshold (e.g. `|dy|>th_trans`) — they should surface as a distinct interval.\n7) **Interval compression**: merge consecutive frames with identical label *sets* (order-insensitive) into `\"start->end\"` half-open ranges.\n8) **Fallback**: if the transform estimate fails, use identity → `\"Stay\"`; never emit an empty label list.\n# Decision sketch\n```python\nlabels=[]\nfor i in range(1, len(sample_ids)):\n    M, _ = cv2.estimateAffinePartial2D(prev_pts, curr_pts, method=cv2.RANSAC)\n    if M is None: labels.append([\"Stay\"]); continue\n    dx, dy = M[0,2], M[1,2]\n    a, b = M[0,0], M[1,0]\n    scale = math.hypot(a,b); rot = math.atan2(b,a)\n    lbl=[]\n    if abs(scale-1)>th_scale: lbl.append(\"Dolly In\" if scale>1 else \"Dolly Out\")\n    if abs(rot)>th_rot:       lbl.append(\"Roll Right\" if rot>0 else \"Roll Left\")\n    if abs(dx)>th_trans:      lbl.append(\"Pan Left\"  if dx>0 else \"Pan Right\")\n    if abs(dy)>th_trans:      lbl.append(\"Tilt Up\"   if dy>0 else \"Tilt Down\")\n    if not lbl: lbl.append(\"Stay\")\n    labels.append(lbl)\n```\n# Self-check\n- [ ] Fallback to identity transform on failure; never emit empty labels.\n- [ ] Sign convention explicitly matches CAMERA motion (image-right feature shift ⇒ Pan Left, etc.), and was sanity-checked against one obvious cue in the clip.\n- [ ] Pan and Tilt tested independently — no `|dx|>=|dy|` gating.\n- [ ] Frames whose translation crosses threshold are not silently merged away.\n- [ ] Compressed intervals cover all sampled frames as a half-open partition of `[0, n_samples)`.\n",{"name":5,"description":7},"main","cdcb6d230f3b15c3c33f8a2bdd16439f43602cbe","evaluation\u002Fresults\u002FMMG2Skill-Opus47-Gold15-20260924\u002Ftasks\u002Fdynamic-object-aware-egomotion\u002Frepaired_skill\u002Fegomotion-estimation",1,"https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark","https:\u002F\u002Fgithub.com\u002Flinyouxue","2026-11-14 20:28:35",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add linyouxue\u002Fskill-repair-benchmark --skill egomotion-estimation",{"tool":37,"command":38},"bunx","bunx skills add linyouxue\u002Fskill-repair-benchmark --skill egomotion-estimation",{"tool":40,"command":41},"pnpm","pnpm dlx skills add linyouxue\u002Fskill-repair-benchmark --skill egomotion-estimation","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fevaluation\u002Fresults\u002FMMG2Skill-Opus47-Gold15-20260924\u002Ftasks\u002Fdynamic-object-aware-egomotion\u002Frepaired_skill\u002Fegomotion-estimation\u002FSKILL.md",true,[],[46,53,60,67,72,77],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":19},"01M7E1HGVRY73NBJ022CJG13X6","video-frame-extraction","Extract frames from video files and save them as images using OpenCV","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_03\u002Finputs\u002Fskills\u002Fvideo-frame-extraction\u002FSKILL.md",[],[],{"skillNo":54,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":19},"01M7E1HAG5R7CJ9C8NYJS771MB","openai-vision","Analyze images and multi-frame sequences using OpenAI GPT vision models","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_03\u002Finputs\u002Fskills\u002Fopenai-vision\u002FSKILL.md",[],[],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":19},"01M7E1H7DADKB6QBWKC2KQ63SZ","image-ocr","Extract text content from images using Tesseract OCR via Python, including structured field extraction from receipts and forms","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_03\u002Finputs\u002Fskills\u002Fimage-ocr\u002FSKILL.md",[],[],{"skillNo":68,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":19},"01M7E1H0WBNT43TENGQY7AVZHD","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_02\u002Finputs\u002Fskills\u002Fvideo-frame-extraction\u002FSKILL.md",[],[],{"skillNo":73,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7E1GTC7D4N4Q2EF03875J9E","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_02\u002Finputs\u002Fskills\u002Fopenai-vision\u002FSKILL.md",[],[],{"skillNo":78,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7E1GQ92CWFMGH53KSSE4WER","https:\u002F\u002Fgithub.com\u002Flinyouxue\u002Fskill-repair-benchmark\u002Fblob\u002Fmain\u002Fjpg-ocr-stat\u002Fresult\u002Fskilltta-claude4.7\u002Fgroup_03\u002Frun_02\u002Finputs\u002Fskills\u002Fimage-ocr\u002FSKILL.md",[],[]]