MR-GDINO：高效的开放世界持续目标检测
MR-GDINO: Efficient Open-World Continual Object Detection
AI总结：
针对开放世界持续目标检测中未见类别的灾难性遗忘问题，提出基于记忆与检索机制的MR-GDINO基线，仅激活0.1%额外参数即大幅缓解遗忘，在旧、新及未见类别上均达到最先进性能。
摘要：
开放世界（OW）识别与检测模型展现出强大的零样本与少样本适应能力，启发了将其作为持续学习方法的初始化以提升性能。尽管在已见类别上取得了可喜的结果，但由于灾难性遗忘，这种在未见类别上的OW能力大幅退化。为应对这一挑战，我们提出了一项开放世界持续目标检测任务，要求检测器在持续学习场景中泛化至旧、新及未见类别。基于此任务，我们提出了一个具挑战性且实用的OW-COD基准来评估检测能力。其目标是促使OW检测器在少样本适应下同时保留已学类别、适应新类别并维持开放世界能力。为缓解未见类别的遗忘，我们提出了MR-GDINO，这是一个通过高度可扩展记忆池内的记忆与检索机制构建的强大、高效且可扩展的基线。实验结果表明，现有持续检测器在已见和未见类别上均存在严重遗忘。相比之下，MR-GDINO仅激活0.1%的额外参数便大幅缓解了遗忘，在旧、新及未见类别上均实现了最先进的性能。
英文摘要：
Open-world (OW) recognition and detection models show strong zero- and few-shot adaptation abilities, inspiring their use as initializations in continual learning methods to improve performance. Despite promising results on seen classes, such OW abilities on unseen classes are largely degenerated due to catastrophic forgetting. To tackle this challenge, we propose an open-world continual object detection task, requiring detectors to generalize to old, new, and unseen categories in continual learning scenarios. Based on this task, we present a challenging yet practical OW-COD benchmark to assess detection abilities. The goal is to motivate OW detectors to simultaneously preserve learned classes, adapt to new classes, and maintain open-world capabilities under few-shot adaptations. To mitigate forgetting in unseen categories, we propose MR-GDINO, a strong, efficient and scalable baseline via memory and retrieval mechanisms within a highly scalable memory pool. Experimental results show that existing continual detectors suffer from severe forgetting for both seen and unseen categories. In contrast, MR-GDINO largely mitigates forgetting with only 0.1% activated extra parameters, achieving state-of-the-art performance for old, new, and unseen categories.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
