MemFlow：面向快速域自适应特征映射的轻量前向记忆框架
MemFlow: A Lightweight Forward Memorizing Framework for Quick Domain Adaptive Feature Mapping
用冻结骨干网络加随机连接脉冲神经元的前向记忆机制替代反向传播微调，为视觉模型在边缘设备上做在线域自适应提供了一条无需梯度的快速路径，可直接用于测试时自适应场景。
原文依据 a lightweight gradient-free forward-memorizing framework, namely MemFlow, which leverages a frozen backbone
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论在真实环境中部署预训练视觉模型时，由于测试场景的多样性，模型性能往往会出现显著下降。通过从目标域采集的无标签数据在边缘设备上对学习模型进行持续自适应，是提升泛化能力的有效手段。然而，基于梯度反向传播对深度神经网络中海量参数进行优化，其耗时远高于前向推理，使得在线学习在低功耗边缘设备上难以实现。为应对这一关键挑战，我们提出了一个轻量的无梯度前向记忆框架 MemFlow，它利用冻结的骨干网络，能够高效地微调特征与预测之间的映射关系。具体而言，MemFlow 采用随机连接的神经元来记忆特征与标签之间的关联；在网络内部传播脉冲信号，并根据置信度水平关联神经元所存储的记忆来生成预测。更值得注意的是，MemFlow 支持利用无标签数据对特征映射进行强化记忆，从而能够快速自适应到新的域。在四个真实世界跨域数据集上的大量实验表明，MemFlow 在仅消耗传统域自适应方法所需计算时间不到 1% 的情况下，性能提升最高可达 10%。
英文原文摘要 1430 字符
Deploying pretrained visual models in real-world environments often suffers from significant performance degradation due to the diversity of testing scenarios. Continuous adaptation of learning models on edge devices via unlabeled data collected from the target domain is highly effective for boosting generalization capability. However, gradient-backpropagation-based optimization of the massive parameters in deep neural networks is vastly more time-consuming than forward inference, rendering online learning infeasible on low-power edge devices. To address this critical challenge, we propose a lightweight gradient-free forward-memorizing framework, namely MemFlow, which leverages a frozen backbone and enables efficient fine-tuning of the mapping between features and predictions. Specifically, MemFlow employs randomly connected neurons to memorize feature-label associations; within the network, spiking signals are propagated, and predictions are generated by associating neuron-stored memories according to their confidence levels. More notably, MemFlow supports reinforced memorization of feature mappings using unlabeled data, thereby enabling rapid adaptation to new domains. Extensive experiments on four real-world cross-domain datasets demonstrate that MemFlow achieves performance improvements of up to 10% while consuming less than 1% of the computational time required by traditional domain adaptation methods.