Instant-Teaching: An End-to-End Semi-Supervised Object Detection Framework

CVPR 2021  ·  Qiang Zhou, Chaohui Yu, Zhibin Wang, Qi Qian, Hao Li ·

Supervised learning based object detection frameworks demand plenty of laborious manual annotations, which may not be practical in real applications. Semi-supervised object detection (SSOD) can effectively leverage unlabeled data to improve the model performance, which is of great significance for the application of object detection models. In this paper, we revisit SSOD and propose Instant-Teaching, a completely end-to-end and effective SSOD framework, which uses instant pseudo labeling with extended weak-strong data augmentations for teaching during each training iteration. To alleviate the confirmation bias problem and improve the quality of pseudo annotations, we further propose a co-rectify scheme based on Instant-Teaching, denoted as Instant-Teaching$^*$. Extensive experiments on both MS-COCO and PASCAL VOC datasets substantiate the superiority of our framework. Specifically, our method surpasses state-of-the-art methods by 4.2 mAP on MS-COCO when using $2\%$ labeled data. Even with full supervised information of MS-COCO, the proposed method still outperforms state-of-the-art methods by about 1.0 mAP. On PASCAL VOC, we can achieve more than 5 mAP improvement by applying VOC07 as labeled data and VOC12 as unlabeled data.

PDF Abstract CVPR 2021 PDF CVPR 2021 Abstract

Results from the Paper

Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Semi-Supervised Object Detection COCO 100% labeled data Instant-Teaching mAP 40.2 # 12
Semi-Supervised Object Detection COCO 10% labeled data Instant Teaching mAP 30.40 # 24
Semi-Supervised Object Detection COCO 1% labeled data Instant Teaching mAP 18.05 # 20
Semi-Supervised Object Detection COCO 2% labeled data Instant Teaching mAP 22.45 # 15
Semi-Supervised Object Detection COCO 5% labeled data Instant Teaching mAP 26.75 # 21


No methods listed for this paper. Add relevant methods here