Hands Deep in Deep Learning for Hand Pose Estimation
This work addresses hand pose estimation for applications like human-computer interaction, but it is incremental as it builds on existing CNN architectures with specific enhancements.
The paper tackles 3D hand pose estimation from depth maps by introducing a prior on 3D pose and using context to resolve finger ambiguities, resulting in significant improvements in accuracy and computation times over state-of-the-art methods on challenging benchmarks.
We introduce and evaluate several architectures for Convolutional Neural Networks to predict the 3D joint locations of a hand given a depth map. We first show that a prior on the 3D pose can be easily introduced and significantly improves the accuracy and reliability of the predictions. We also show how to use context efficiently to deal with ambiguities between fingers. These two contributions allow us to significantly outperform the state-of-the-art on several challenging benchmarks, both in terms of accuracy and computation times.