Multiple instance active learning for object detection
Despite the substantial progress of active learning for image recognition, there still lacks an instance-level active learning method specified for object detection. In this paper, we propose Multiple Instance Active Object Detection (MI-AOD), to select the most informative images for detector training by observing instance-level uncertainty. MI-AOD defines an instance uncertainty learning module, which leverages the discrepancy of two adversarial instance classifiers trained on the labeled set to predict instance uncertainty of the unlabeled set. MI-AOD treats unlabeled images as instance bags and feature anchors in images as instances, and estimates the image uncertainty by re-weighting instances in a multiple instance learning (MIL) fashion. Iterative instance uncertainty learning and re-weighting facilitate suppressing noisy instances, toward bridging the gap between instance uncertainty and image-level uncertainty. Experiments validate that MI-AOD sets a solid baseline for instance-level active learning. On commonly used object detection datasets, MI-AOD outperforms state-of-the-art methods with significant margins, particularly when the labeled sets are small. Code is available at https://github.com/yuantn/MI-AOD.
PDF Abstract CVPR 2021 PDF CVPR 2021 AbstractCode
Datasets
Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
---|---|---|---|---|---|---|
Active Object Detection | MS COCO | RetinaNet | AP | (7.3, 13.8, 16.9, 19.1, 20.8) on 2% ~ 10% | # 1 | |
Active Object Detection | PASCAL VOC 07+12 | RetinaNet | mAP | (47.18, 58.41, 64.02, 67.72, 69.79, 71.07, 72.27) on 5% ~ 20% | # 1 | |
Active Object Detection | PASCAL VOC 07+12 | SSD | mAP | (53.62, 62.86, 66.83, 69.33, 70.80, 72.21, 72.84, 73.74, 74.18, 74.91) on 1k ~ 10k | # 1 |