Negative Prototypes Guided Contrastive Learning for WSOD
This addresses the challenge of object detection without bounding box annotations for researchers in computer vision, though it appears incremental as it builds on existing contrastive learning paradigms.
The paper tackles the problem of Weakly Supervised Object Detection (WSOD) with only image-level annotations by proposing Negative Prototypes Guided Contrastive Learning (NPGC), which leverages negative prototypes and a global feature bank to improve feature representation, achieving state-of-the-art performance on VOC07 and VOC12 datasets.
Weakly Supervised Object Detection (WSOD) with only image-level annotation has recently attracted wide attention. Many existing methods ignore the inter-image relationship of instances which share similar characteristics while can certainly be determined not to belong to the same category. Therefore, in order to make full use of the weak label, we propose the Negative Prototypes Guided Contrastive learning (NPGC) architecture. Firstly, we define Negative Prototype as the proposal with the highest confidence score misclassified for the category that does not appear in the label. Unlike other methods that only utilize category positive feature, we construct an online updated global feature bank to store both positive prototypes and negative prototypes. Meanwhile, we propose a pseudo label sampling module to mine reliable instances and discard the easily misclassified instances based on the feature similarity with corresponding prototypes in global feature bank. Finally, we follow the contrastive learning paradigm to optimize the proposal's feature representation by attracting same class samples closer and pushing different class samples away in the embedding space. Extensive experiments have been conducted on VOC07, VOC12 datasets, which shows that our proposed method achieves the state-of-the-art performance.