RNNPool: Efficient Non-linear Pooling for RAM Constrained Inference

TL;DR

RNNPool introduces a recurrent neural network-based pooling method that reduces memory and computation in CNNs, enabling efficient inference on resource-constrained devices without sacrificing accuracy.

Contribution

It proposes RNNPool, a novel non-linear pooling operator that efficiently aggregates features and downsamples activation maps, suitable for edge devices.

Findings

RNNPool effectively replaces multiple layers in CNN architectures.

It significantly reduces memory and computational requirements.

Achieves state-of-the-art face detection on microcontrollers.

Abstract

Standard Convolutional Neural Networks (CNNs) designed for computer vision tasks tend to have large intermediate activation maps. These require large working memory and are thus unsuitable for deployment on resource-constrained devices typically used for inference on the edge. Aggressively downsampling the images via pooling or strided convolutions can address the problem but leads to a significant decrease in accuracy due to gross aggregation of the feature map by standard pooling operators. In this paper, we introduce RNNPool, a novel pooling operator based on Recurrent Neural Networks (RNNs), that efficiently aggregates features over large patches of an image and rapidly downsamples activation maps. Empirical evaluation indicates that an RNNPool layer can effectively replace multiple blocks in a variety of architectures such as MobileNets, DenseNet when applied to standard vision tasks like image classification and face detection. That is, RNNPool can significantly decrease computational complexity and peak memory usage for inference while retaining comparable accuracy. We use RNNPool with the standard S3FD architecture to construct a face detection method that achieves state-of-the-art MAP for tiny ARM Cortex-M4 class microcontrollers with under 256 KB of RAM. Code is released at https://github.com/Microsoft/EdgeML.