English

Optimizing Hand Region Detection in MediaPipe Holistic Full-Body Pose Estimation to Improve Accuracy and Avoid Downstream Errors

Computer Vision and Pattern Recognition 2024-05-14 v2

Abstract

This paper addresses a critical flaw in MediaPipe Holistic's hand Region of Interest (ROI) prediction, which struggles with non-ideal hand orientations, affecting sign language recognition accuracy. We propose a data-driven approach to enhance ROI estimation, leveraging an enriched feature set including additional hand keypoints and the z-dimension. Our results demonstrate better estimates, with higher Intersection-over-Union compared to the current method. Our code and optimizations are available at https://github.com/sign-language-processing/mediapipe-hand-crop-fix.

Cite

@article{arxiv.2405.03545,
  title  = {Optimizing Hand Region Detection in MediaPipe Holistic Full-Body Pose Estimation to Improve Accuracy and Avoid Downstream Errors},
  author = {Amit Moryossef},
  journal= {arXiv preprint arXiv:2405.03545},
  year   = {2024}
}
R2 v1 2026-06-28T16:18:12.182Z