Semi-Supervised Object Detection with Adaptive Class-Rebalancing Self-Training

07/11/2021
by   Fangyuan Zhang, et al.
0

This study delves into semi-supervised object detection (SSOD) to improve detector performance with additional unlabeled data. State-of-the-art SSOD performance has been achieved recently by self-training, in which training supervision consists of ground truths and pseudo-labels. In current studies, we observe that class imbalance in SSOD severely impedes the effectiveness of self-training. To address the class imbalance, we propose adaptive class-rebalancing self-training (ACRST) with a novel memory module called CropBank. ACRST adaptively rebalances the training data with foreground instances extracted from the CropBank, thereby alleviating the class imbalance. Owing to the high complexity of detection tasks, we observe that both self-training and data-rebalancing suffer from noisy pseudo-labels in SSOD. Therefore, we propose a novel two-stage filtering algorithm to generate accurate pseudo-labels. Our method achieves satisfactory improvements on MS-COCO and VOC benchmarks. When using only 1% labeled data in MS-COCO, our method achieves 17.02 mAP improvement over supervised baselines, and 5.32 mAP improvement compared with state-of-the-art methods.

READ FULL TEXT
research
02/18/2021

Unbiased Teacher for Semi-Supervised Object Detection

Semi-supervised learning, i.e., training networks with both labeled and ...
research
12/08/2020

3DIoUMatch: Leveraging IoU Prediction for Semi-Supervised 3D Object Detection

3D object detection is an important yet demanding task that heavily reli...
research
04/09/2022

S4OD: Semi-Supervised learning for Single-Stage Object Detection

Single-stage detectors suffer from extreme foreground-background class i...
research
03/23/2022

Scale-Equivalent Distillation for Semi-Supervised Object Detection

Recent Semi-Supervised Object Detection (SS-OD) methods are mainly based...
research
06/08/2021

Salvage of Supervision in Weakly Supervised Detection

Weakly supervised object detection (WSOD) has recently attracted much at...
research
10/17/2022

Dual-Curriculum Teacher for Domain-Inconsistent Object Detection in Autonomous Driving

Object detection for autonomous vehicles has received increasing attenti...
research
12/10/2021

Label, Verify, Correct: A Simple Few Shot Object Detection Method

The objective of this paper is few-shot object detection (FSOD) – the ta...

Please sign up or login with your details

Forgot password? Click here to reset