WS-iFSD: Weakly Supervised Incremental Few-shot Object Detection Without Forgetting

Xinyu Gong, Li Yin, Juan-Manuel Perez-Rua, Zhangyang Wang, Zhicheng Yan

Keywords: few-shot object detection

TL;DR: Our Incremental Few-Shot Object Detection (iFSD) framework uses meta-learning and Weakly Supervised Class Augmentation to detect objects in both base and novel categories, significantly outperforming state-of-the-art methods on multiple benchmarks.

Abstract: Traditional object detection algorithms rely on extensive annotations from a pre-defined set of base categories, leaving them ill-equipped to identify objects from novel classes. We address this limitation by introducing a novel framework for Incremental Few-Shot Object Detection (iFSD). Leveraging a meta-learning approach, our \hypernetwork is designed to generate class-specific codes, enabling object recognition from both base and novel categories. To enhance the \hypernetwork's generalization performance, we propose a Weakly Supervised Class Augmentation technique that significantly amplifies the training data by merely requiring image-level labels for object localization. Additionally, we stabilize detection performance on base categories by freezing the backbone and detection heads during meta-training. Our model demonstrates significant performance gains on two major benchmarks. Specifically, it outperforms the state-of-the-art ONCE approach on the MS COCO dataset by margins of $2.8\%$ and $20.5\%$ in box AP for novel and base categories, respectively. When trained on MS COCO and cross-evaluated on PASCAL VOC, our model achieves a four-fold improvement in box AP compared to ONCE.

Track Confirmation: Yes, I am submitting to the proceeding track.

Submission Number: 36