This works builds on the matrix backpropagation methodology introduced in our previous paper: Matrix Backpropagation for Deep Networks with Structured Layers, C. Ionescu and O. Vantzos and C. Sminchisescu.
The following six papers were accepted to CVPR18:
- Monocular 3D Pose and Shape Estimation of Multiple People in Natural Scenes - The Importance of Multiple Scene Constraints, A. Zanfir, E. Marinoiu, C. Sminchisescu
- 3D Human Sensing, Action and Emotion Recognition in Robot Assisted Therapy of Children With Autism, E. Marinoiu, M. Zanfir, V. Olaru, C. Sminchisescu
- Deep Learning of Graph Matching, A. Zanfir, C. Sminchisescu
- Human Appearance Transfer, M. Zanfir, A. Popa, A. Zanfir, C. Sminchisescu
- Semantic Video Segmentation by Gated Recurrent Flow Propagation, D. Nilsson, C. Sminchisescu
- Deep Reinforcement Learning of Region Proposal Networks for Object Detection, A. Pirinen, C. Sminchisescu
Check our videos on how well people can re-enact the 3D pose of peers from photographs!
Talk available here.
Code available here.
See our structured prediction and mixture of experts methods. Code available. See also the IJCV 2010 Special Issue: Evaluation of Articulated Human Motion and Pose Estimation.