Bridge Past and Future: Overcoming Information Asymmetry in Incremental Object Detection
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914872538693632 |
|---|---|
| author | Mo, Qijie Gao, Yipeng Fu, Shenghao Yan, Junkai Wu, Ancong Zheng, Wei-Shi |
| author_facet | Mo, Qijie Gao, Yipeng Fu, Shenghao Yan, Junkai Wu, Ancong Zheng, Wei-Shi |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_11499 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Bridge Past and Future: Overcoming Information Asymmetry in Incremental Object Detection Mo, Qijie Gao, Yipeng Fu, Shenghao Yan, Junkai Wu, Ancong Zheng, Wei-Shi Computer Vision and Pattern Recognition 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. |
| title | Bridge Past and Future: Overcoming Information Asymmetry in Incremental Object Detection |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2407.11499 |