Solving Missing-Annotation Object Detection with Background Recalibration Loss
This addresses a challenging problem in computer vision for researchers and practitioners dealing with incomplete annotations, though it is incremental as it builds on existing loss formulations like Focal Loss.
The paper tackles object detection when most true objects are unlabeled in datasets, causing missing-labeled areas to be treated as background during training, and introduces Background Recalibration Loss (BRL) to automatically recalibrate loss signals, achieving superior performance over baselines and state-of-the-art methods on PASCAL VOC and MS COCO datasets.
This paper focuses on a novel and challenging detection scenario: A majority of true objects/instances is unlabeled in the datasets, so these missing-labeled areas will be regarded as the background during training. Previous art on this problem has proposed to use soft sampling to re-weight the gradients of RoIs based on the overlaps with positive instances, while their method is mainly based on the two-stage detector (i.e. Faster RCNN) which is more robust and friendly for the missing label scenario. In this paper, we introduce a superior solution called Background Recalibration Loss (BRL) that can automatically re-calibrate the loss signals according to the pre-defined IoU threshold and input image. Our design is built on the one-stage detector which is faster and lighter. Inspired by the Focal Loss formulation, we make several significant modifications to fit on the missing-annotation circumstance. We conduct extensive experiments on the curated PASCAL VOC and MS COCO datasets. The results demonstrate that our proposed method outperforms the baseline and other state-of-the-arts by a large margin. Code available: https://github.com/Dwrety/mmdetection-selective-iou.