AI Motion Capture
Ọrụ ọhaneze OK
380+ Models
Enweghị akara mmiri
Enweghị mkpa ịbanye
Móòdù:
+ GPT-5, Claude, Gemini
Wepụta vidiyo ọ bụla na onye na ya - AI na-enyocha 33 body keypoints kwa fréèmù ma na-enye gị skeleton overlay video nakwa JSON nke mpaghara ebe maka frèèmù ọ bụla. Enweghị mocap suit, enweghị akara, enweghị calibration. Single-camera markerless motion capture site na MediaPipe.
Na-enyocha 33 body joints n'elu fréèmù ọbụla...
Nkarachọ ndị ahụ
Ihenhọrọ ahụ
Token na-aga n'okpuru.
Wepụta Token
❤️ Ị hụrụ Free.ai? Kpọtụrụ enyi gị!
Tinye aka ka ị nweta njikọ n'aka enyi gị ma nweta 25,000 token kwa enyi.
Ịchọrọ ihe ọzọ?
Sign up free maka 30K tokens / ụbọchị + 10K bonus
Akaụntụ
Na-arụ ọrụ n'ihe ịchọrọ...
Wepụta vidiyo, AI na-ewepụ 3D body pose kwa fréèmù site na iji MediaPipe. Zụta azụ skeleton overlay vidiyo nakwa per-fréèmù keypoints JSON maka animation, sports analysis, ma ọ bụ biomechanics. Free, enweghị akara, enweghị mocap suit.
Etu esi eji ya AI Motion Capture
1
Tinye init gị
Tinye ngwe, bubata faịlụ, mọọbụ depụta ihe ịchọrọ. Achọrọ akaụntụ ọbụla.
2
Pịa mepụta
Anyị AI na-arụ ọrụ gị n'ime sekọnd na-eji ihe kacha mma open-source models.
3
Bubata na akwado
Bubata, debata, mọọbụ kesaa nsonaazụ gị. Free maka ojiji nkeonwe na nke azụmahịa.
Jiri tùlè a site na API
Megharịa ihenhọrọ a site na koodị gị. OpenAI-compatible REST endpoint, Bearer-token auth, enweghị SDK ọzọ achọrọ. Token costs match the web interface.
curl -X POST https://api.free.ai/v1/video/generate/ \
-H "Authorization: Bearer sk-free-..." \
-H "Content-Type: application/json" \
-d '{"prompt": "A cat playing piano", "duration": 4}'AI Motion Capture — FAQ
N'ime vidiyo na onye na-eme ihe na-eme ihe na AI na-enyocha 33 nke ahụ - isi, ụkwụ, elbows, wrists, hips, knee, ankles, nakwa aka na ụkwụ - na freịm ọ bụla. I nwere ike ịmaliteghachi vidiyo skeleton overlay nakwa faịlụ JSON na-arụ ọrụ na-arụ ọrụ na-arụ ọrụ. Enweghị mocap suit, enweghị akara, enweghị nzọụkwụ calibration, otu igwefoto na-arụ ọrụ dị mma.
Kwụsị 3D animations nke ihenhọrọ na Blender / Unity / Unreal (na-ezigaghachi n'ime armalite nke na-adọba ụgbọala), mee egwuregwu / egwu / nchọpụta usoro iheomume, mepụta mbido-ntọlite ntọlite, zụlite ML model na data mmegharị, ma ọ bụ lelee mmegharị ntọlite n'oge.
MediaPipe Pose Landmarker (Google, Apache 2.0). Ọ na-eweta 33 body keypoints kwa fréèmù na-echekwa 2D coords + a na-atụle Z (relative depth site na kamọnèra) + per-keypoint visibility scores. Ọ na-arụ ọrụ kpamkpam na CPU ka GPU na-adị n'efu maka ndị ọzọ gị.
Ọ bụ 2.5D - eziokwu 2D + ekwenyeghị Z site na igwefoto otu. N'ezie 3D motion capture chọrọ ọtụtụ igwefoto synchronized maka triangulation (nke FreeMoCap / OptiTrack / Vicon approach). Maka TikTok dance, analysis egwuregwu, animation reference, ma ọ bụ ọ bụla single-camera workflow, MediaPipe's output bụ ezi uche dị na ya. Anyị ga-agbakwunye multi-camera tool mgbe ahụ maka ndị ọrụ chọrọ eziokwu 3D.
200 tokens kwa sekọnd nke input video (floored na 500 tokens). 10-sekọnd clip na-efu 2,000 tokens; 60-sekọnd clip na-efu 12,000. Daily-pool free tokens na-ekpuchi ụfọdụ n'ime clip n'ụbọchị; ndị ọrụ banye-na-enweta 30K/ụbọchị.
Real-time-ish: ihe dị ka 30-50 frame kwa sekọnd na igbe anyị. A 1-minute 30 fps video usoro na 30-60 sekọnd end-to-end gụnyere ibudata + render. Longer videos take proportionally longer.
MP4, MOV, WebM, AVI, MKV, nakwa ọtụtụ vidio fóráịtị ndị ọzọ — ihe ọbụla ffmpeg nwere ike iwepụ ụghalaahịa. Max upload 100 MB. Nhazi ahụ abụghị ihe dị mkpa; ụdị n'ime ya na-ewepụtụla maka ọsọ.
MediaPipe Pose Landmarker na-enyocha onye ONE n'ime fréèmù (onye kasị n'anya). Maka ndị mmadụ-ọtụtụ-enyocha anyị ga-achọ ụdị dị iche iche (RTMPose, YOLOv8-pose). Ọ bụrụ na ihenhọrọ gị bụ ndị mmadụ-ọtụtụ, tinye echiche site na /contact/ — ọ ga-enwe obi ụtọ ịgbakwunye ya dịka otu ihenhọrọ.
Ọrụ ndị na-ahụ anya a na-enyocha ha n'ime ~5-10 px na 720p fréèmù; ọrụ ndị na-ekpuchi (aka n'azụ azụ, aka n'ebe ọha-fréèmù) na-abịa na-apụtaghị n'ihu n'ihi ya ị nwere ike ịsacha ha. Nhazi n'etiti fréèmù na n'okpuru pipeline gị (Kalman / Savitzky-Golay) na-ewepụ ihe ndị ọzọ.
Ọ bụghị site n'otu ụzọ n'otu na tùleè a taa - JSON bụ kiiboodu data na-enweghị isi. I nwere ike ịgbanwe n'ọfụụ site n'iji laịibụlụ dịka `aniposelib` mọọbụ `pose-format`. Anyị na-atụle ịnye "Mocap → BVH/FBX" na-eso ụzọ tùleè - tinye upvote na /contact/ ọ bụrụ na ịchọrọ ya.
E mepụtara ya n'oge na-adịghị anya, kiiboodu ndị ahụ a na-ewepụ, mgbe ahụ a na-ehichapụ input video. Skeleton-overlay output na JSON a na-echekwa maka standard share-link expiry (24 h anonymous / 7 d paid). A na-ejighị ya maka nkuzi. /privacy/ maka ntọala zuru ezu.
Ya - POST multipart `video` faịlụ na /v1/video/motion-capture/. Na-eziga {video_url, json_url, duration_s, tokens, share_url}. Okwenye onye na-eweta (sk-free-…) na-enye gị 10,000 tokens/ọnwa n'efu. Curl n'ụdị na /api/.
Olee otú ị ga-esi họrọ ihenhọrọ a?