Subtensor Quantization for Mobilenets
This work addresses accuracy degradation in quantized Mobilenets for efficient deployment, but it is incremental as it builds on existing quantization methods.
The paper tackled the problem of quantization loss in Mobilenet architectures by analyzing root causes and proposing alternatives, achieving post-training quantized 8-bit inference with top-1 accuracy within 0.7% of the floating point version on ImageNet.
Quantization for deep neural networks (DNN) have enabled developers to deploy models with less memory and more efficient low-power inference. However, not all DNN designs are friendly to quantization. For example, the popular Mobilenet architecture has been tuned to reduce parameter size and computational latency with separable depth-wise convolutions, but not all quantization algorithms work well and the accuracy can suffer against its float point versions. In this paper, we analyzed several root causes of quantization loss and proposed alternatives that do not rely on per-channel or training-aware approaches. We evaluate the image classification task on ImageNet dataset, and our post-training quantized 8-bit inference top-1 accuracy in within 0.7% of the floating point version.