Ukuthatha umsindo we-AI
Ukusetshenziswa kwebhizinisi OK
Amamodeli angama-380+
Akukho phawu lwamanzi
Akukho ubhaliso olungenayo
Imodeli:
+ GPT-5, Claude, Gemini
Layisha phezulu noma iyiphi ividiyo enemuntu kuyo — i-AI ilandelela ama-keypoints we-33 we-body ngefreyimu futhi ikunikeza i-skeleton overlay video kanye ne-JSON yezindawo ezihlangene zefreyimu ngayinye. Akukho mocap suit, akukho ama-markers, akukho ukukala. Ikhamera eyodwa engabonisi ukuqapha kwe-motion nge-MediaPipe.
Ukulandelela 33 umzimba izixhumi phezu ngayinye ifreyimu…
Izinketho ezithuthukisiwe
Imiphumela
Ama-token aphansi.
Thola ama-token amaningi
❤️ Uthanda i-Free.ai? Ngisho nabahlobo bakho!
Ngena ukuze uthole i-link yokuxhumanisa futhi uthole ama-token angama-25,000 ngamunye umngane.
Ufuna okuningi?
Bhala mahhala nge-30K tokens / ngosuku + 10K ibhonasi
Bhala
Kuqedwa isicelo sakho...
Layisha ividiyo, i-AI ikhipha i-3D body pose ngefreyimu usebenzisa iMediaPipe. Thola i-skeleton overlay video kanye ne-per-frame keypoints JSON ye-animations, ucwaningo lwezemidlalo, noma i-biomechanics. Imahhala, akukho ama-markers, akukho mocap suit.
Indlela Yokusetshenziswa Ukuthatha umsindo we-AI
1
Faka umngenisi wakho
Bhala umbhalo, ulayishe ifayela, noma uchaza ofuna ukukwenza. Akukho akhawunti edingekayo.
2
Chofoza ukwakha
I-AI yethu isebenza isicelo sakho emaminithini usebenzisa amamodeli avulekile angcono kakhulu.
3
Layisha phezulu futhi uhlukanise
Layisha phezulu, kopela, noma uhlukanise imiphumela yakho. Imahhala ukusetshenziswa komuntu siqu nokwebhizinisi.
Sebenzisa lesi sisetshenziswa nge-API
Yenza lokhu sixhobo sisebenze ngokuzenzakalela kusuka kukhodi yakho. I-REST ehambisanayo ne-OpenAI, ukuqinisekisa i-token ye-bearer, akukho SDK edingekayo. Izindleko ze-token zifana ne-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}'Ukuthatha umsindo we-AI — FAQ
Drop in a video with a person in frame and the AI tracks 33 body joints — head, shoulders, elbows, wrists, hips, knees, ankles, plus hands and feet — in every frame. You get back a skeleton overlay video plus a JSON file with the per-frame joint coordinates. No mocap suit, no markers, no calibration step, single camera works fine.
Uqhube izithombe eziphilayo zesikhashana zesikhashana ze-3D ku-Blender / Unity / Unreal (uphinde ubeke inhloso ku-armature efakwe i-rigged), yenza ucwaningo lwezithako / umdanso / i-martial-arts technique, kwakha ifomu lokulungisa izingqimba, qeqesha amamodeli we-ML ku-movement data, noma ubone kuphela imikhuba yokushintshana ngesikhathi.
I-MediaPipe Pose Landmarker (Google, Apache 2.0). Ikhipha ama-keypoints we-33 we-body nge-frame nge-2D coords + i-Z elinganiselwe (ububanzi obuphathelene nekhamera) + ama-keypoints abonakalayo nge-score. Isebenza ngokuphelele ku-CPU ngakho-ke i-GPU ihlala ikhululekile kwezinye izizukulwane zakho.
I-2.5D — 2D eqiniso + i-Z elinganiselwe ephathelene nekhamera eyodwa. Ukuqapha kwe-3D kwe-real 3D kudinga amakhamera axhunywe ngokuzenzakalela ahlukahlukene ukuhlukanisa ama-triangulation (uhlelo lwe-FreeMoCap / OptiTrack / Vicon). Ukudlalwa kwe-TikTok, ukuhlaziywa kwezemidlalo, ukubhekisa kwe-animations, noma noma iyiphi i-camera-i
Ama-token angama-200 ngomzuzu ngamunye wevidiyo yokungena (i-floored ku-500 tokens). I-clip engu-10 ngomzuzu ibiza ama-token angama-2,000; i-clip engu-60 ngomzuzu ibiza ama-token angama-12,000. Ama-token amahhala e-pool agcwele ama-clip ambalwa aphansi ngosuku; abasebenzisi ababhalisiwe bathola ama-30K/day.
Isikhathi sangempela: cishe amafreyimu angama-30-50 ngesekondi kwibhokisi lethu. Imizuzu engu-1, ama-30 fps amavidiyo asebenza ngama-30-60 sekondi kusukela ekuqaleni kuya ekugcineni kufaka phakathi uku upload + ukuveza. Amavidiyo ade kakhulu athatha isikhathi eside.
MP4, MOV, WebM, AVI, MKV, kanye nefomethi yevidiyo ejwayelekile kakhulu — noma yini i-ffmpeg ingayifaka. Ukukhuphula okuphezulu 100 MB. Ukuxazulula akuyona into ebalulekile; imodeli ye-pose iphansi ngokwengxenye ukukhawulela.
I-MediaPipe Pose Landmarker ilandelela umuntu ONYE ngefreyimu (obalulekile kakhulu). Ukulandelana kwe-multi-person sidinga imodeli ehlukile (RTMPose, YOLOv8-pose). Uma kusetshenziswa i-multi-person, thumela imibono nge /contact/ — siyavuya ukuyifaka njengethuluzi elilodwa.
Izinhlanganisela ezibonakalayo zilandelela ngaphakathi ~5-10 px kwifreyimu ye-720p; izinhlanganisela ezivalayo (isandla ngasemva, ingalo ngaphandle kwefreyimu) zigcwele ngezinga lokubona okuphansi ukuze ukwazi ukuzihlunga. Ukugcoba phakathi kwefreyimu ku-downstream pipeline yakho (Kalman / Savitzky-Golay) kuhlanza konke okusele.
Akunakho okuqondile kusuka kulolu thuluzi namuhla — i-JSON iyidatha yendawo ebalulekile engaqondakali. Ungaguqula ngaphandle kwe-intanethi usebenzisa ama-library njenge-`aniposelib` noma i-`pose-format`. Sicabanga ukuthutha i-"Mocap → BVH/FBX" elandelanayo — faka i-upvote ku /contact/ uma ufuna.
Iqhubekelwe ngokushesha, amaphuzu aphambili akhishwa, bese ividiyo yokungena isuswa. I-skeleton-overlay output ne-JSON zigcinwa ngesilinganiso sokuphelelwa isikhathi sokuxhumana (24 h angaziwayo / 7 d akhokhelwe). Akufanele kusetshenziswa ukuqeqeshwa. /privacy/ ngenqubo egcwele.
Yebo — POST ihele le-`video` elihlukanisiwe libe yi /v1/video/motion-capture/. Libuyisela {video_url, json_url, duration_s, tokens, share_url}. Ugunyaziso lwesisebenzi (sk-free-…) lunikeza 10,000 tokens/ngonyaka mahhala. Umzekelo we-curl ku /api/.
Ungayilinganisela kanjani lethuluzi?