FORGE：微控制器整数视觉模型的前向测试时自适应
原标题：FORGE: Forward-Only Test-Time Adaptation for Integer-Only Vision Models on Microcontrollers
AI 摘要
FORGE 提出了一种面向微控制器上整数部署视觉模型的前向测试时自适应方法。该方法通过重新归一化折叠卷积的逐通道输出，恢复了因批归一化折叠而丢失的统计信息，从而在无反向传播的推理运行时实现自适应。实验表明，FORGE 在 CIFAR-10/100 和 Tiny-ImageNet 上恢复了 TENT 的大部分精度提升，并在 ESP32-S3 上实测自适应能耗仅为推理能量的 6.8%。
正文节选
FORGE: Forward-Only Test-Time Adaptation for Integer-Only Vision Models on Microcontrollers Abstract Vision models deployed on microcontrollers (MCUs) are quantized to integer-only arithmetic and run in inference-only runtimes that do not carry the machinery backpropagation needs: the standard tool for adapting a model to the distribution shift (sensor noise, blur, lighting) it meets in the field. Existing forward-only test-time adaptation (TTA) methods either run only on server- or edge-GPU-cla