Over the past few years, detection performance improvements of deep-learning based steganalyzers have been usually achieved through structure expansion. However, excessive expanded structure results in huge computational cost, storage overheads, and consequently difficulty in training and deployment. In this paper we propose CALPA-NET, a ChAnneL-Pruning-Assisted deep residual network architecture search approach to shrink the network structure of existing vast, over-parameterized deep-learning based steganalyzers. We observe that the broad inverted-pyramid structure of existing deep-learning based steganalyzers might contradict the well-established model diversity oriented philosophy, and therefore is not suitable for steganalysis. Then a hybrid criterion combined with two network pruning schemes is introduced to adaptively shrink every involved convolutional layer in a data-driven manner. The resulting network architecture presents a slender bottleneck-like structure. We have conducted extensive experiments on BOSSBase + BOWS2 dataset, more diverse ALASKA dataset and even a large-scale subset extracted from ImageNet CLS-LOC dataset. The experimental results show that the model structure generated by our proposed CALPA-NET can achieve comparative performance with less than two percent of parameters and about one third FLOPs compared to the original steganalytic model. The new model possesses even better adaptivity, transferability, and scalability.

中文翻译：

---

CALPA-NET：用于数字图像隐写分析的通道修剪辅助深度残差网络


过去几年，基于深度学习的隐写分析器的检测性能提升通常是通过结构扩展来实现的。然而，过度扩展的结构会导致巨大的计算成本、存储开销，从而导致训练和部署困难。在本文中，我们提出了 CALPA-NET，这是一种 ChAnneL 剪枝辅助深度残差网络架构搜索方法，用于缩小现有庞大的、基于超参数化深度学习的隐写分析器的网络结构。我们观察到，现有基于深度学习的隐写分析器的广泛倒金字塔结构可能与完善的模型多样性导向哲学相矛盾，因此不适合隐写分析。然后引入与两种网络修剪方案相结合的混合标准，以数据驱动的方式自适应地缩小每个涉及的卷积层。由此产生的网络架构呈现出细长的瓶颈状结构。我们对 BOSSBase + BOWS2 数据集、更多样化的 ALASKA 数据集甚至从 ImageNet CLS-LOC 数据集提取的大规模子集进行了广泛的实验。实验结果表明，与原始隐写分析模型相比，我们提出的 CALPA-NET 生成的模型结构可以用不到百分之二的参数和大约三分之一的 FLOP 来实现比较性能。新模型具有更好的适应性、可移植性和可扩展性。