Bridge Past and Future: Overcoming Information Asymmetry in Incremental Object Detection
This addresses a practical limitation in incremental learning for object detection where co-occurring objects from different stages create inconsistent optimization, with incremental improvements to existing methods.
The paper tackles catastrophic forgetting in incremental object detection by addressing information asymmetry when images contain objects from past, present, and future stages, proposing a method (BPF) that aligns models across stages and introduces a novel distillation loss; it achieves state-of-the-art performance on Pascal VOC and MS COCO benchmarks without requiring memory.
In incremental object detection, knowledge distillation has been proven to be an effective way to alleviate catastrophic forgetting. However, previous works focused on preserving the knowledge of old models, ignoring that images could simultaneously contain categories from past, present, and future stages. The co-occurrence of objects makes the optimization objectives inconsistent across different stages since the definition for foreground objects differs across various stages, which limits the model's performance greatly. To overcome this problem, we propose a method called ``Bridge Past and Future'' (BPF), which aligns models across stages, ensuring consistent optimization directions. In addition, we propose a novel Distillation with Future (DwF) loss, fully leveraging the background probability to mitigate the forgetting of old classes while ensuring a high level of adaptability in learning new classes. Extensive experiments are conducted on both Pascal VOC and MS COCO benchmarks. Without memory, BPF outperforms current state-of-the-art methods under various settings. The code is available at https://github.com/iSEE-Laboratory/BPF.