@oliviscusAI: this tool can track perfect 3D motion. rtmlib is a lightweight pose estimation library covering full body, hands, face,…

X AI KOLs Following Tools

Summary

rtmlib is a lightweight, open-source pose estimation library that supports full-body, hand, face, and animal pose tracking, built on rtmpose and vitpose models, with a built-in Gradio web UI.

this tool can track perfect 3D motion. rtmlib is a lightweight pose estimation library covering full body, hands, face, and 133-point wholebody tracking, built on rtmpose and vitpose models. > needs only numpy, opencv, and onnxruntime to run > covers 30 named animal species alongside human pose > includes a built-in gradio webui, one script to launch it > runs on cpu, gpu, or openvino, models auto-download on first use 100% free. open source.
Original Article
View Cached Full Text

Cached at: 09/15/26, 01:34 AM

this tool can track perfect 3D motion.

rtmlib is a lightweight pose estimation library covering full body, hands, face, and 133-point wholebody tracking, built on rtmpose and vitpose models.

needs only numpy, opencv, and onnxruntime to run covers 30 named animal species alongside human pose includes a built-in gradio webui, one script to launch it runs on cpu, gpu, or openvino, models auto-download on first use

100% free. open source.

Similar Articles

MolmoMotion: Language-guided 3D motion forecasting

Hugging Face Blog

MolmoMotion is a new language-guided 3D motion forecasting model that predicts future 3D point trajectories from video frames and action descriptions, achieving stronger performance than existing methods. Alongside the model, a large dataset (MolmoMotion-1M) and a benchmark (PointMotionBench) are released.

DreamTraj: Generating 6-DoF Object Trajectories by Reading Unrendered Video Diffusion Latents

Hugging Face Daily Papers

DreamTraj predicts 6-DoF object trajectories from a single RGB image and a language instruction by decoding internal video diffusion latents, eliminating the need for video, depth, or CAD models at inference. It introduces the MOVEdataset with fine-grained language-to-motion annotations and achieves state-of-the-art performance while running 4.6x faster than generate-then-extract pipelines.