Depth-Guided Self-Supervised Human Keypoint Detection via Cross-Modal Distillation

TL;DR

This paper introduces Distill-DKP, a self-supervised framework that uses depth maps to improve human keypoint detection, significantly outperforming previous methods by leveraging cross-modal knowledge distillation.

Contribution

The paper presents a novel cross-modal knowledge distillation approach that incorporates depth information to enhance unsupervised human keypoint detection.

Findings

Reduces mean L2 error by 47.15% on Human3.6M

Improves keypoint accuracy by 1.3% on DeepFashion

Demonstrates effective knowledge transfer across network layers

Abstract

Existing unsupervised keypoint detection methods apply artificial deformations to images such as masking a significant portion of images and using reconstruction of original image as a learning objective to detect keypoints. However, this approach lacks depth information in the image and often detects keypoints on the background. To address this, we propose Distill-DKP, a novel cross-modal knowledge distillation framework that leverages depth maps and RGB images for keypoint detection in a self-supervised setting. During training, Distill-DKP extracts embedding-level knowledge from a depth-based teacher model to guide an image-based student model with inference restricted to the student. Experiments show that Distill-DKP significantly outperforms previous unsupervised methods by reducing mean L2 error by 47.15% on Human3.6M, mean average error by 5.67% on Taichi, and improving keypoints accuracy by 1.3% on DeepFashion dataset. Detailed ablation studies demonstrate the sensitivity of knowledge distillation across different layers of the network. Project Page: https://23wm13.github.io/distill-dkp/