View-Invariant Probabilistic Embedding for Human Pose
arXiv:1912.01001
Abstract
Depictions of similar human body configurations can vary with changing viewpoints. Using only 2D information, we would like to enable vision algorithms to recognize similarity in human body poses across multiple views. This ability is useful for analyzing body movements and human behaviors in images and videos. In this paper, we propose an approach for learning a compact view-invariant embedding space from 2D joint keypoints alone, without explicitly predicting 3D poses. Since 2D poses are projected from 3D space, they have an inherent ambiguity, which is difficult to represent through a deterministic mapping. Hence, we use probabilistic embeddings to model this input uncertainty. Experimental results show that our embedding model achieves higher accuracy when retrieving similar poses across different camera views, in comparison with 2D-to-3D pose lifting models. We also demonstrate the effectiveness of applying our embeddings to view-invariant action recognition and video alignment. Our code is available at https://github.com/google-research/google-research/tree/master/poem.
arXiv commentAccepted to ECCV 2020 (Spotlight presentation). Code is available at https://github.com/google-research/google-research/tree/master/poem . Video synchronization results are available at https://drive.google.com/corp/drive/folders/1kTc_UT0Eq0H2ZBgfEoh8qEJMFBouC-Wv
References in corpus (8)
- Auto-Encoding Variational Bayes
- What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
- In Defense of the Triplet Loss for Person Re-Identification
- Word Representations via Gaussian Embedding
- Multi-task Deep Learning for Real-Time 3D Human Pose Estimation and Action Recognition
- Unsupervised Learning of View-invariant Action Representations
- Modeling Uncertainty with Hedged Instance Embedding
- Pose Embeddings: A Deep Architecture for Learning to Match Human Poses