DeepPrior++: Improving Fast and Accurate 3D Hand Pose Estimation

TL;DR

DeepPrior++ enhances 3D hand pose estimation from depth maps by integrating ResNet, data augmentation, and improved localization, achieving state-of-the-art results with a simple approach.

Contribution

It introduces simple yet effective improvements to DeepPrior, significantly boosting performance while maintaining simplicity.

Findings

Outperforms recent methods on NYU, ICVL, MSRA benchmarks

Achieves comparable or better accuracy with simpler model

Open-source implementation available

Abstract

DeepPrior is a simple approach based on Deep Learning that predicts the joint 3D locations of a hand given a depth map. Since its publication early 2015, it has been outperformed by several impressive works. Here we show that with simple improvements: adding ResNet layers, data augmentation, and better initial hand localization, we achieve better or similar performance than more sophisticated recent methods on the three main benchmarks (NYU, ICVL, MSRA) while keeping the simplicity of the original method. Our new implementation is available at https://github.com/moberweger/deep-prior-pp .