Saved in:
Bibliographic Details
Main Authors: Ai, Yihao, Qi, Yifei, Wang, Bo, Cheng, Yu, Wang, Xinchao, Tan, Robby T.
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2407.15451
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913441538637824
author Ai, Yihao
Qi, Yifei
Wang, Bo
Cheng, Yu
Wang, Xinchao
Tan, Robby T.
author_facet Ai, Yihao
Qi, Yifei
Wang, Bo
Cheng, Yu
Wang, Xinchao
Tan, Robby T.
contents Existing 2D human pose estimation research predominantly concentrates on well-lit scenarios, with limited exploration of poor lighting conditions, which are a prevalent aspect of daily life. Recent studies on low-light pose estimation require the use of paired well-lit and low-light images with ground truths for training, which are impractical due to the inherent challenges associated with annotation on low-light images. To this end, we introduce a novel approach that eliminates the need for low-light ground truths. Our primary novelty lies in leveraging two complementary-teacher networks to generate more reliable pseudo labels, enabling our model achieves competitive performance on extremely low-light images without the need for training with low-light ground truths. Our framework consists of two stages. In the first stage, our model is trained on well-lit data with low-light augmentations. In the second stage, we propose a dual-teacher framework to utilize the unlabeled low-light data, where a center-based main teacher produces the pseudo labels for relatively visible cases, while a keypoints-based complementary teacher focuses on producing the pseudo labels for the missed persons of the main teacher. With the pseudo labels from both teachers, we propose a person-specific low-light augmentation to challenge a student model in training to outperform the teachers. Experimental results on real low-light dataset (ExLPose-OCN) show, our method achieves 6.8% (2.4 AP) improvement over the state-of-the-art (SOTA) method, despite no low-light ground-truth data is used in our approach, in contrast to the SOTA method. Our code will be available at:https://github.com/ayh015-dev/DA-LLPose.
format Preprint
id arxiv_https___arxiv_org_abs_2407_15451
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Domain-Adaptive 2D Human Pose Estimation via Dual Teachers in Extremely Low-Light Conditions
Ai, Yihao
Qi, Yifei
Wang, Bo
Cheng, Yu
Wang, Xinchao
Tan, Robby T.
Computer Vision and Pattern Recognition
Existing 2D human pose estimation research predominantly concentrates on well-lit scenarios, with limited exploration of poor lighting conditions, which are a prevalent aspect of daily life. Recent studies on low-light pose estimation require the use of paired well-lit and low-light images with ground truths for training, which are impractical due to the inherent challenges associated with annotation on low-light images. To this end, we introduce a novel approach that eliminates the need for low-light ground truths. Our primary novelty lies in leveraging two complementary-teacher networks to generate more reliable pseudo labels, enabling our model achieves competitive performance on extremely low-light images without the need for training with low-light ground truths. Our framework consists of two stages. In the first stage, our model is trained on well-lit data with low-light augmentations. In the second stage, we propose a dual-teacher framework to utilize the unlabeled low-light data, where a center-based main teacher produces the pseudo labels for relatively visible cases, while a keypoints-based complementary teacher focuses on producing the pseudo labels for the missed persons of the main teacher. With the pseudo labels from both teachers, we propose a person-specific low-light augmentation to challenge a student model in training to outperform the teachers. Experimental results on real low-light dataset (ExLPose-OCN) show, our method achieves 6.8% (2.4 AP) improvement over the state-of-the-art (SOTA) method, despite no low-light ground-truth data is used in our approach, in contrast to the SOTA method. Our code will be available at:https://github.com/ayh015-dev/DA-LLPose.
title Domain-Adaptive 2D Human Pose Estimation via Dual Teachers in Extremely Low-Light Conditions
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2407.15451