Projection Convolutional Neural Networks for 1-bit CNNs via Discrete Back Propagation
This work addresses the challenge of deploying deep convolutional neural networks in resource-constrained settings, offering an incremental improvement in compression and accuracy for binarized networks.
The paper tackles the problem of improving binarized neural networks (BNNs) for resource-constrained environments by introducing projection convolutional neural networks (PCNNs) with discrete back propagation via projection, achieving the best classification performance on ImageNet and CIFAR datasets compared to other state-of-the-art BNNs.
The advancement of deep convolutional neural networks (DCNNs) has driven significant improvement in the accuracy of recognition systems for many computer vision tasks. However, their practical applications are often restricted in resource-constrained environments. In this paper, we introduce projection convolutional neural networks (PCNNs) with a discrete back propagation via projection (DBPP) to improve the performance of binarized neural networks (BNNs). The contributions of our paper include: 1) for the first time, the projection function is exploited to efficiently solve the discrete back propagation problem, which leads to a new highly compressed CNNs (termed PCNNs); 2) by exploiting multiple projections, we learn a set of diverse quantized kernels that compress the full-precision kernels in a more efficient way than those proposed previously; 3) PCNNs achieve the best classification performance compared to other state-of-the-art BNNs on the ImageNet and CIFAR datasets.