Progressive Learning of Low-Precision Networks
This addresses the challenge of deploying efficient deep learning models on resource-limited platforms like mobile devices, though it is incremental as it builds on existing low-precision network methods.
The paper tackles the problem of training low-precision neural networks, which often suffer from accuracy degradation, by proposing a progressive learning strategy that expands networks during training and removes expansions for inference, resulting in faster convergence and higher accuracy on datasets like SVHN, CIFAR, and ILSVRC-2012.
Recent years have witnessed the great advance of deep learning in a variety of vision tasks. Many state-of-the-art deep neural networks suffer from large size and high complexity, which makes it difficult to deploy in resource-limited platforms such as mobile devices. To this end, low-precision neural networks are widely studied which quantize weights or activations into the low-bit format. Though being efficient, low-precision networks are usually hard to train and encounter severe accuracy degradation. In this paper, we propose a new training strategy through expanding low-precision networks during training and removing the expanded parts for network inference. First, we equip each low-precision convolutional layer with an ancillary full-precision convolutional layer based on a low-precision network structure, which could guide the network to good local minima. Second, a decay method is introduced to reduce the output of the added full-precision convolution gradually, which keeps the resulted topology structure the same to the original low-precision one. Experiments on SVHN, CIFAR and ILSVRC-2012 datasets prove that the proposed method can bring faster convergence and higher accuracy for low-precision neural networks.