3D Human Motion Estimation via Motion Compression and Refinement
This work addresses the challenge of 3D human motion estimation for computer vision applications, but it appears incremental as it builds on existing auto-encoder techniques.
The paper tackles the problem of generating smooth and accurate 3D human pose and motion estimates from RGB video sequences, achieving results that are both smooth and accurate.
We develop a technique for generating smooth and accurate 3D human pose and motion estimates from RGB video sequences. Our method, which we call Motion Estimation via Variational Autoencoder (MEVA), decomposes a temporal sequence of human motion into a smooth motion representation using auto-encoder-based motion compression and a residual representation learned through motion refinement. This two-step encoding of human motion captures human motion in two stages: a general human motion estimation step that captures the coarse overall motion, and a residual estimation that adds back person-specific motion details. Experiments show that our method produces both smooth and accurate 3D human pose and motion estimates.