Adaptive Pruning of Channel Spatial Dependability in Convolutional Neural Networks

Weiying Xie, Mei Yuan, Jitao Ma, Yunsong Li

Abstract: Deep Convolutional Neural Networks (CNNs) have demonstrated excellent performance in various multimedia application scenarios. However, complex models often require significant computational resources and energy costs. Therefore, CNN compression is crucial for addressing deployment challenges of multimedia application on resource constrained edge devices. However, existing CNN channel pruning strategies primarily focus on the "weights" or "activations" of the model, overlooking its "interpretability" information. In this paper, we explore CNN pruning strategies from the perspective of model interpretability. We model the correspondence between channel feature maps and interpretable visual perception based on class saliency maps, aiming to assess the contribution of each channel to the desired output. Additionally, we utilize Discrete Wavelet Transform (DWT) to capture the global features and structure of class saliency maps. Based on this, we propose a Channel Spatial Dependability (CSD) metric, evaluating the importance and contribution of channels in a bidirectional manner to guide model quantization pruning. And we dynamically adjust the pruning rate of each layer based on performance changes, in order to achieve more accurate and efficient adaptive pruning. Experimental results demonstrate that our method achieves significant results across a range of different networks and datasets. For instance, we achieved a 51.3% pruning on the ResNet-56 model while maintaining an accuracy of 94.16%, outperforming feature-map or weight-based pruning and other State-of-the-Art (SOTA).

Primary Subject Area: [Experience] Multimedia Applications

Relevance To Conference: Multimedia processing typically involves a large amount of data and complex models, such as image recognition, video processing, etc. By pruning, the size and computational complexity of the model can be effectively reduced, thereby improving processing speed and efficiency. For some multimedia processing applications that require deployment on edge devices, storage and bandwidth resources are very valuable. Pruning can reduce the storage requirements of the model, thereby saving storage space and reducing the bandwidth required for the model during transmission.

Submission Number: 3877