SASL：用于神经网络加速的显著性自适应稀疏学习
SASL: Saliency-Adaptive Sparsity Learning for Neural Network Acceleration
摘要：
加快CNN的推理速度对于它们在现实世界应用中的部署至关重要。在所有修剪方法中，那些实现稀疏学习框架的方法已经证明是有效的，因为它们以端到端数据驱动的方式学习和修剪模型。然而，这些工作强加相同的稀疏正则化的所有过滤器不加区别，这很难导致最佳的结构稀疏网络。在本文中，我们提出了一个显着性自适应稀疏学习（SASL）的方法进行进一步优化。每个滤波器的新颖且有效的估计，即，显着性，这是衡量从两个方面：预测性能的重要性和消耗的计算资源。在稀疏学习过程中，正则化强度根据显着性进行调整，因此我们的优化格式可以更好地保持预测性能，同时将更多计算量较大的过滤器归零。显著性的计算为训练过程引入了最小的开销，这意味着我们的SASL非常高效。在剪枝阶段，为了优化所提出的数据依赖准则，采用了硬样本挖掘策略，该策略具有更高的有效性和效率。大量的实验证明了我们的方法的优越性能。值得注意的是，在ILSVRC-2012数据集上，我们的方法可以减少ResNet-50的49.7%的FLOP，而top-1和top-5的准确度下降可以忽略不计。
英文摘要：
Accelerating the inference speed of CNNs is critical to their deployment in real-world applications. Among all the pruning approaches, those implementing a sparsity learning framework have shown to be effective as they learn and prune the models in an end-to-end data-driven manner. However, these works impose the same sparsity regularization on all filters indiscriminately, which can hardly result in an optimal structure-sparse network. In this paper, we propose a Saliency-Adaptive Sparsity Learning (SASL) approach for further optimization. A novel and effective estimation of each filter, i.e., saliency, is designed, which is measured from two aspects: the importance for the prediction performance and the consumed computational resources. During sparsity learning, the regularization strength is adjusted according to the saliency, so our optimized format can better preserve the prediction performance while zeroing out more computation-heavy filters. The calculation for saliency introduces minimum overhead to the training process, which means our SASL is very efficient. During the pruning phase, in order to optimize the proposed data-dependent criterion, a hard sample mining strategy is utilized, which shows higher effectiveness and efficiency. Extensive experiments demonstrate the superior performance of our method. Notably, on ILSVRC-2012 dataset, our approach can reduce 49.7% FLOPs of ResNet-50 with very negligible 0.39% top-1 and 0.05% top-5 accuracy degradation.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
