在持续全景分割中寻求平衡
Strike a Balance in Continual Panoptic Segmentation
AI总结：
本研究针对持续全景分割任务，提出了BalConpas方法，通过过去类回溯蒸馏、类比例记忆策略和平衡反误导损失，在稳定性与适应性、重放样本效用及不完整标注处理间实现平衡，在ADE20K数据集上取得SOTA性能。
摘要：
本研究探索了持续全景分割这一新兴领域，强调了三个关键的平衡。首先，我们引入了过去类回溯蒸馏，以平衡现有知识的稳定性和对新信息的适应性。该技术根据最终的标签分配结果，回溯与过去类相关的特征，并从之前的模型中针对这些特定特征执行知识蒸馏，同时允许其他特征灵活地适应新信息。此外，我们引入了一种类比例记忆策略，将重放样本集中的类分布与历史训练数据的类分布对齐。该策略在重放过程中保持了平衡的类表示，增强了有限容量重放样本集在回忆先前类时的效用。此外，认识到重放样本仅针对其原始步骤的类进行标注，我们设计了平衡的反误导损失，在不引起分类偏差的情况下对抗不完整标注的影响。基于这些创新，我们提出了一种名为BalConpas（Balanced Continual Panoptic Segmentation）的新方法。我们在具有挑战性的ADE20K数据集上的评估表明，与现有的最先进方法相比，其性能更优。官方代码可在https://github.com/jinpeng0528/BalConpas获取。
英文摘要：
This study explores the emerging area of continual panoptic segmentation, highlighting three key balances. First, we introduce past-class backtrace distillation to balance the stability of existing knowledge with the adaptability to new information. This technique retraces the features associated with past classes based on the final label assignment results, performing knowledge distillation targeting these specific features from the previous model while allowing other features to flexibly adapt to new information. Additionally, we introduce a class-proportional memory strategy, which aligns the class distribution in the replay sample set with that of the historical training data. This strategy maintains a balanced class representation during replay, enhancing the utility of the limited-capacity replay sample set in recalling prior classes. Moreover, recognizing that replay samples are annotated only for the classes of their original step, we devise balanced anti-misguidance losses, which combat the impact of incomplete annotations without incurring classification bias. Building upon these innovations, we present a new method named Balanced Continual Panoptic Segmentation (BalConpas). Our evaluation on the challenging ADE20K dataset demonstrates its superior performance compared to existing state-of-the-art methods. The official code is available at https://github.com/jinpeng0528/BalConpas.
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。
