动态调整模型适应能力，缓解新旧知识冲突，提升持续学习效果。
BPG: Balancing Plasticity and Generalization for Domain Incremental Learning
- 根据数据特征可分性动态调整适配器维度，实现灵活容量分配。
- 测试时软融合多领域模型，使准确率提升至93.1%，遗忘率低至0.22%。
- 适用于需要长期更新的视觉识别场景，尤其适合资源受限部署。
深度神经网络在各类任务中表现优异，但在数据分布持续变化时泛化能力下降，导致性能显著退化。领域增量学习（DIL）旨在让模型持续适应新领域的同时保留已有知识。现有基于参数隔离的方法虽达当前最优，但普遍采用统一方式处理各新领域，造成学习能力不足或参数冗余。本文提出BPG统一框架，包含两个互补组件：BPG-Adapter根据领域特征可分性动态决定适配器隐藏层维度；BPG-Inference采用软领域混合策略，在测试时融合多个领域专用模型，缓解领域标识误选问题。在DomainNet、CDDB和CORe50上的实验表明，BPG持续优于固定适配器方法与硬选择策略，在DomainNet上遗忘率低至0.22%，平均准确率达到93.1%，达到当前最优水平。
原文摘要 · Abstract (English)
Deep neural networks excel in various tasks but struggle to generalize across evolving data distributions, leading to significant performance degradation under domain shifts. Domain incremental learning (DIL) addresses this challenge by enabling models to continuously adapt while retaining prior knowledge. Among existing DIL approaches, the parameter-isolation paradigm achieves state-of-the-art performance. However, these methods often adopt a one-size-fits-all approach to adapt to new domains, resulting in either insufficient learning capacity or redundant parameters. In this work, we propose BPG, a unified framework that addresses both challenges through two complementary components: BPG-Adapter, which dynamically determines each domain's adapter hidden dimension based on domain-specific feature separability, and BPG-Inference, a soft domain mixture strategy that integrates multiple domain-specific models at test time, mitigating domain ID misselection. Experimental results on DomainNet, CDDB, and CORe50 demonstrate that BPG consistently outperforms uniform adapter-based approaches and hard domain selection strategies, achieving state-of-the-art average accuracy while reducing forgetting to as low as 0.22% on DomainNet.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。