基于媒体分类任务的可解释多尺度信息块的定位
Localizing Interpretable Multi-scale informative Patches Derived from Media Classification Task
摘要：
深度卷积神经网络（CNN）总是依赖于更宽的感受野（RF）和更复杂的非线性来实现最先进的性能，同时也越来越难以解释相关补丁如何贡献最终预测。在本文中，我们构建了一个可解释的锚网配备了我们精心设计的RF和线性空间聚合，以提供逐块的输入媒体的可解释性，同时本地化的多尺度信息补丁只监督媒体级别的标签，没有任何额外的限界框注释。局部信息图像和文本块的可视化显示了AnchorNet优越的多尺度定位能力。我们进一步使用本地化补丁在广泛应用的网络中进行下游分类任务。实验结果表明，将原始输入替换为局部化后的分类块，在性能损失很小的情况下，能够获得明显的推理加速效果，证明局部化后的分类块确实能够保留原始输入的大部分语义和证据。
英文摘要：
Deep convolutional neural networks (CNN) always depend on wider receptive field (RF) and more complex non-linearity to achieve state-of-the-art performance, while suffering the increased difficult to interpret how relevant patches contribute the final prediction. In this paper, we construct an interpretable AnchorNet equipped with our carefully designed RFs and linearly spatial aggregation to provide patch-wise interpretability of the input media meanwhile localizing multi-scale informative patches only supervised on media-level labels without any extra bounding box annotations. Visualization of localized informative image and text patches show the superior multi-scale localization capability of AnchorNet. We further use localized patches for downstream classification tasks across widely applied networks. Experimental results demonstrate that replacing the original inputs with their patches for classification can get a clear inference acceleration with only tiny performance degradation, which proves that localized patches can indeed retain the most semantics and evidences of the original inputs.
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。
