Structured Probabilistic Pruning for Convolutional Neural Network Acceleration
This addresses the need for efficient neural network deployment, particularly in resource-constrained environments, by offering a novel pruning approach that is incremental but with strong performance improvements.
The paper tackles the problem of accelerating convolutional neural networks by proposing Structured Probabilistic Pruning (SPP), a method that prunes weights probabilistically instead of deterministically, resulting in a 4x speedup for AlexNet with only 0.3% top-5 accuracy loss and similar gains for VGG-16 and ResNet-50.
In this paper, we propose a novel progressive parameter pruning method for Convolutional Neural Network acceleration, named Structured Probabilistic Pruning (SPP), which effectively prunes weights of convolutional layers in a probabilistic manner. Unlike existing deterministic pruning approaches, where unimportant weights are permanently eliminated, SPP introduces a pruning probability for each weight, and pruning is guided by sampling from the pruning probabilities. A mechanism is designed to increase and decrease pruning probabilities based on importance criteria in the training process. Experiments show that, with 4x speedup, SPP can accelerate AlexNet with only 0.3% loss of top-5 accuracy and VGG-16 with 0.8% loss of top-5 accuracy in ImageNet classification. Moreover, SPP can be directly applied to accelerate multi-branch CNN networks, such as ResNet, without specific adaptations. Our 2x speedup ResNet-50 only suffers 0.8% loss of top-5 accuracy on ImageNet. We further show the effectiveness of SPP on transfer learning tasks.