1st Place Solutions for UG2+ Challenge 2021 -- (Semi-)supervised Face detection in the low light condition

TL;DR

This paper presents a top-performing semi-supervised face detection method for low-light conditions, combining image enhancement, transfer techniques, and multiple detection frameworks to achieve state-of-the-art results.

Contribution

The authors developed a novel ensemble approach integrating image enhancement and transfer methods with various detection frameworks for improved low-light face detection.

Findings

Achieved 74.89 mAP on the test set

Ensemble of multiple models outperforms individual models

Effective use of image transfer methods improves detection performance

Abstract

In this technical report, we briefly introduce the solution of our team "TAL-ai" for (Semi-) supervised Face detection in the low light condition in UG2+ Challenge in CVPR 2021. By conducting several experiments with popular image enhancement methods and image transfer methods, we pulled the low light image and the normal image to a more closer domain. And it is observed that using these data to training can achieve better performance. We also adapt several popular object detection frameworks, e.g., DetectoRS, Cascade-RCNN, and large backbone like Swin-transformer. Finally, we ensemble several models which achieved mAP 74.89 on the testing set, ranking 1st on the final leaderboard.