在无排练记忆的情况下利用任务特定令牌减少增量学习中的灾难性遗忘
Reducing catastrophic forgetting of incremental learning in the absence of rehearsal memory with task-specific token
AI总结：
提出一种基于视觉变换器架构、利用任务特定令牌封装任务知识的方法，在不存储旧数据的情况下减少增量学习中的灾难性遗忘，并在基准测试中取得最优准确率和最低反向迁移。
摘要：
深度学习模型在持续学习新数据时通常会表现出灾难性遗忘。许多增量学习方法通过在学习新任务时重用先前任务的数据来解决这一问题。然而，直接访问过去的数据会引发隐私和安全问题。为了解决这些问题，我们提出了一种新颖的方法，在不存储先前数据的情况下保留先前的知识。该方法受视觉变换器架构的启发，采用一种独特的令牌，能够封装每个任务的压缩知识。该方法通过根据与数据关联的任务以不同方式引导注意力来生成任务特定的嵌入，从而有效地通过令牌模拟拥有多个模型的影响。我们的方法结合了蒸馏过程，确保即使在多次额外的学习步骤之后也能进行高效的交互，从而优化模型以防止遗忘。我们使用基准数据集在不同任务增量学习场景下，以准确率和反向迁移来衡量我们模型的性能。我们的结果表明了我们方法的优越性，在比较的方法中实现了最高的准确率和最低的反向迁移。除了提出新模型外，我们的方法还为视觉变换器架构范围内的各种扩展奠定了基础。
英文摘要：
Deep learning models generally display catastrophic forgetting when learning new data continuously. Many incremental learning approaches address this problem by reusing data from previous tasks while learning new tasks. However, the direct access to past data generates privacy and security concerns. To address these issues, we present a novel method that preserves previous knowledge without storing previous data. This method is inspired by the architecture of a vision transformer and employs a unique token capable of encapsulating the compressed knowledge of each task. This approach generates task-specific embeddings by directing attention differently based on the task associated with the data, thereby effectively mimicking the impact of having multiple models through tokens. Our method incorporates a distillation process that ensures efficient interactions even after multiple additional learning steps, thereby optimizing the model against forgetting. We measured the performance of our model in terms of accuracy and backward transfer using a benchmark dataset for different task-incremental learning scenarios. Our results demonstrate the superiority of our approach, which achieved the highest accuracy and lowest backward transfer among the compared methods. In addition to presenting a new model, our approach lays the foundation for various extensions within the spectrum of vision-transformer architectures.
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。
