PGTRNet: Two-phase Weakly Supervised Object Detection with Pseudo Ground Truth Refinement
This work addresses the problem of improving object detection accuracy with weak supervision for computer vision researchers, representing an incremental advancement in the field.
The paper tackles performance issues in two-phase weakly supervised object detection by proposing PGTRNet, which uses multiple bounding boxes and online pseudo ground truth refinement to improve learning and decouple models, achieving a 2.1% mAP boost and state-of-the-art results on PASCAL VOC 2007.
Current state-of-the-art weakly supervised object detection (WSOD) studies mainly follow a two-stage training strategy which integrates a fully supervised detector (FSD) with a pure WSOD model. There are two main problems hindering the performance of the two-phase WSOD approaches, i.e., insufficient learning problem and strict reliance between the FSD and the pseudo ground truth (PGT) generated by the WSOD model. This paper proposes pseudo ground truth refinement network (PGTRNet), a simple yet effective method without introducing any extra learnable parameters, to cope with these problems. PGTRNet utilizes multiple bounding boxes to establish the PGT, mitigating the insufficient learning problem. Besides, we propose a novel online PGT refinement approach to steadily improve the quality of PGT by fully taking advantage of the power of FSD during the second-phase training, decoupling the first and second-phase models. Elaborate experiments are conducted on the PASCAL VOC 2007 benchmark to verify the effectiveness of our methods. Experimental results demonstrate that PGTRNet boosts the backbone model by 2.1% mAP and achieves the state-of-the-art performance.