Chirality Nets for Human Pose Regression
This work addresses human pose regression tasks, such as 3D pose estimation and activity recognition, by exploiting left/right symmetry, but it is incremental as it builds on existing equivariant methods.
The authors tackled the problem of human pose regression by proposing Chirality Nets, a family of deep nets equivariant to chirality transforms, which improved data efficiency and reduced computation. The approach achieved or matched state-of-the-art results, with more significant gains on small datasets and limited-data settings.
We propose Chirality Nets, a family of deep nets that is equivariant to the "chirality transform," i.e., the transformation to create a chiral pair. Through parameter sharing, odd and even symmetry, we propose and prove variants of standard building blocks of deep nets that satisfy the equivariance property, including fully connected layers, convolutional layers, batch-normalization, and LSTM/GRU cells. The proposed layers lead to a more data efficient representation and a reduction in computation by exploiting symmetry. We evaluate chirality nets on the task of human pose regression, which naturally exploits the left/right mirroring of the human body. We study three pose regression tasks: 3D pose estimation from video, 2D pose forecasting, and skeleton based activity recognition. Our approach achieves/matches state-of-the-art results, with more significant gains on small datasets and limited-data settings.