Epipolar Transformer for Multi-View Human Pose Estimation

Yihui He, Rui Yan, Katerina Fragkiadaki, Shoou-I Yu; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2020, pp. 1036-1037

Abstract


A common way to localize 3D human joints in a synchronized and calibrated multi-view setup is a two-step process: (1) apply a 2D detector separately on each view to localize joints in 2D, and (2) robust triangulation on 2D detections from each view to acquire 3D joint locations. However, in step 1, the 2D detector is constrained to solve challenging cases which could be better resolved in 3D, such as occlusions and oblique viewing angles, purely in 2D without leveraging any 3D information. Therefore, we propose the differentiable "epipolar transformer", which empowers the 2D detector to leverage 3D-aware intermediate features to improve 2D pose estimation. The intuition is: given a 2D location p in reference view, we would like to first find its corresponding point p' in source view, then combine the features at p' with the features at p, thus leading to a more 3D-aware intermediate feature at p. Inspired by stereo matching, the epipolar transformer leverages epipolar constraints and feature matching to approximate the features at p'. The key advantages of the epipolar transformer is: (1) it has minimal learnable parameters, (2) it can be easily plugged into existing networks, moreover (3) it is easily interpretable, i.e., we can analyze the location p' to understand whether matching over the epipolar line was successful. Experiments on InterHand and Human3.6M show that our approach has consistent improvements over the baselines. Specifically, in the condition where no external data is used, our Human3.6M model trained with ResNet-50 and image size 256x256 outperforms state-of-the-art by a large margin and achieves MPJPE 26.9 mm. Code is available. This is the workshop version of our CVPR 2020 paper [8]

Related Material


[pdf]
[bibtex]
@InProceedings{He_2020_CVPR_Workshops,
author = {He, Yihui and Yan, Rui and Fragkiadaki, Katerina and Yu, Shoou-I},
title = {Epipolar Transformer for Multi-View Human Pose Estimation},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {June},
year = {2020}
}