少即是多：剪枝对CNN可解释性的影响
Less is More: The Influence of Pruning on the Explainability of CNNs
AI总结：
该研究探讨网络剪枝带来的模型复杂度降低是否能提升CNN的感知可解释性，通过预研究及两项人类实验测试不同压缩率的影响，发现低压缩率有正向作用，还找到了可同时提升可解释性与性能的平衡点。
摘要：
过去一个世纪，深度学习模型已成为解决复杂计算机视觉问题的最先进方法。这类现代计算机视觉模型拥有数百万个参数，带来了两大主要挑战：（1）计算需求增加，阻碍了其在移动设备或物联网设备等资源受限环境中的部署；（2）难以向人类解释此类网络的复杂决策。网络剪枝是一种降低模型复杂度的技术方法，会移除重要性较低的参数。本文研究了这种技术复杂度的降低是否也有助于提升感知可解释性。为此，我们开展了一项预研究和两项以人为基础的实验，评估不同剪枝比例对可解释性的影响。总体而言，我们在Mechanical Turk平台上通过37500项任务评估了四种不同的压缩率（即2、4、8和32）。结果表明，较低的压缩率对可解释性有积极影响，而较高的压缩率则会产生负面影响。此外，我们还找到了能同时提升感知可解释性和模型性能的最佳平衡点。
英文摘要：
Over the last century, deep learning models have become the state-of-the-art for solving complex computer vision problems. These modern computer vision models have millions of parameters, which presents two major challenges: (1) the increased computational requirements hamper the deployment in resource-constrained environments, such as mobile or IoT devices, and (2) explaining the complex decisions of such networks to humans is challenging. Network pruning is a technical approach to reduce the complexity of models, where less important parameters are removed. The work presented in this paper investigates whether this reduction in technical complexity also helps with perceived explainability. To do so, we conducted a pre-study and two human-grounded experiments, assessing the effects of different pruning ratios on explainability. Overall, we evaluate four different compression rates (i.e., 2, 4, 8, and 32) with 37 500 tasks on Mechanical Turk. Results indicate that lower compression rates have a positive influence on explainability, while higher compression rates show negative effects. Furthermore, we were able to identify sweet spots that increase both the perceived explainability and the model's performance.
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。
