Deep convolutional neural networks have liberated its extraordinary power on various tasks. However, it is still very challenging to deploy state-of-the-art models into real-world applications due to their high computational complexity. How can we design a compact and effective network without massive experiments and expert knowledge? In this paper, we propose a simple and effective framework to learn and prune deep models in an end-to-end manner. In our framework, a new type of parameter -- scaling factor is first introduced to scale the outputs of specific structures, such as neurons, groups or residual blocks. Then we add sparsity regularizations on these factors, and solve this optimization problem by a modified stochastic Accelerated Proximal Gradient (APG) method. By forcing some of the factors to zero, we can safely remove the corresponding structures, thus prune the unimportant parts of a CNN. Compared with other structure selection methods that may need thousands of trials or iterative fine-tuning, our method is trained fully end-to-end in one training pass without bells and whistles. We evaluate our method, Sparse Structure Selection with two state-of-the-art CNNs ResNet and ResNeXt, and demonstrate very promising results with adaptive depth and width selection.

本文提出了一种简单有效的框架来对深度模型进行端到端的剪枝，方法是先引入一个称为“缩放因子”的新参数来缩放特定结构的输出，然后对这些因子加入稀疏正则化，并通过修改的随机加速远端梯度（APG）方法解决这个优化问题。将某些因子强制为零，可以安全地移除对应的结构，从而削减CNN的不重要部分，该方法相较于其他需要数千次试验或迭代微调的结构选择方法具有更好的性能。

基于数据驱动的深度神经网络稀疏结构选择