ã€Unity】OpenPose ==> 3d-pose-baseline-vmd ã§å‡ºåŠ›ã—ãŸé–¢ç¯€åº§æ¨™å€¤ã‚’回転ã«å¤‰æ›ã—ã¦éª¨æ ¼ã‚¢ãƒ‹ãƒ¡ãƒ¼ã‚·ãƒ§ãƒ³ã•ã›ã‚‹ - Qiitaã€Unity】OpenPose ==> 3d-pose-baseline-vmd ã§å‡ºåŠ›ã—ãŸé–¢ç¯€åº§æ¨™å€¤ã‚’回転ã«å¤‰æ›ã—ã¦éª¨æ ¼ã‚¢ãƒ‹ãƒ¡ãƒ¼ã‚·ãƒ§ãƒ³ã•ã›ã‚‹ ã“ã®ãƒ—ãƒã‚°ãƒ©ãƒ を実行ã•ã›ã‚‹ã«ã¯ã€OpenPoseã¨3d-pose-baseline-vmdを使ã£ã¦å‡ºåŠ›ã—ãŸã€pos.txtã¨ã„ã†ãƒ•ã‚¡ã‚¤ãƒ«ãŒå¿…è¦ã«ãªã‚Šã¾ã™ã€‚ 詳ã—ãã¯ã€ https://qiita.com/miu200521358/items/d00c3b6d1bc1b6e67480 ã®è¨˜äº‹ã‚’å‚考ã«ã—ã¦ãã ã•ã„。 pos.txtãŒç”¨æ„ã§æ¥ãŸã‚‰ã€ä¸‹ã®ã‚½ãƒ¼ã‚¹ã‚³ãƒ¼ãƒ‰ã‚’空ã®ã‚²ãƒ¼ãƒ オブジェクトã«ã‚¢ã‚¿ãƒƒãƒã—ã¦ã€ã‚¤ãƒ³ã‚¹ãƒšã‚¯ã‚¿ã«è¡¨ç¤ºã•ã‚Œã‚‹Animatorã«é©å½“ãªã‚ャラクタã®Animatorをドラックã—ã¦è¨å®šã—ã¦ãã ã•ã„。 å‹•ç”»ã§ä½¿ã£ã¦ã„ã‚‹ã‚ャラクタã¯Unitychanã®ã€Œè—¤åŽŸã¿ã•ãã€ã§ã™ã€‚ 時間的余裕ãŒã§ããŸã‚‰ã€å‹•ãã®è€ƒå¯Ÿã‚„ã€ã‚½ãƒ¼ã‚¹ã‚³ãƒ¼ãƒ‰ã®ä¿®æ£ãªã©ã‚’ã—よã†ã¨æ€ã£ã¦ã„
tf-openposeã§3D姿勢推定ã™ã‚‹ã¨ãã«ãƒãƒžã£ãŸè©± - Qiita
GitHub - una-dinosauria/3d-pose-baseline: A simple baseline for 3d human pose estimation in tensorflow. Presented at ICCV 17.
tf-openposeã«ã‚ˆã‚‹å§¿å‹¢æŽ¨å®šã‚’unityã«è½ã¨ã—ã“ã‚‚ã†ã¨ã—ãŸã‘ã©å¤±æ•—ã—ãŸè©± - Qiita
GitHub - zhenyuczy/openpose_3d-pose-baseline_unity3d: Control the 3D model by capturing human joint points with a single camera
GitHub - tensorflow/tfjs-models: Pretrained models for TensorFlow.js
{{#tags}}- {{label}}
{{/tags}}