arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2108.08810cs.CVcs.AIcs.LGstat.ML
收藏

视觉Transformer看起来像卷积神经网络吗?

Do Vision Transformers See Like Convolutional Neural Networks?

Google Research(谷歌研究院)

Maithra Raghu, Thomas Unterthiner, Simon Kornblith, Chiyuan Zhang, Alexey Dosovitskiy

摘要:

到目前为止,卷积神经网络(CNN)已经成为视觉数据的事实模型。最近的工作表明,(视觉)Transformer模型(ViT)可以在图像分类任务上实现相当甚至更高的性能。这就提出了一个核心问题:视觉Transformer是如何解决这些任务的?它们是像卷积网络一样工作,还是学习完全不同的视觉表示?在图像分类基准上分析ViT和CNN的内部表示结构,我们发现这两种架构之间存在显着差异,例如ViT在所有层中具有更统一的表示。我们探讨了这些差异是如何产生的,发现自我注意力发挥了至关重要的作用,这使得全球信息的早期聚合,以及ViT残留连接,这强烈地传播功能从较低到较高层。我们研究了空间定位的后果,证明ViTs成功地保留了输入的空间信息,不同的分类方法有明显的效果。最后,我们研究了(预训练)数据集规模对中间特征和迁移学习的影响,并讨论了与新架构(如MLP混合器)的连接。

英文摘要:

Convolutional neural networks (CNNs) have so far been the de-facto model for visual data. Recent work has shown that (Vision) Transformer models (ViT) can achieve comparable or even superior performance on image classification tasks. This raises a central question: how are Vision Transformers solving these tasks? Are they acting like convolutional networks, or learning entirely different visual representations? Analyzing the internal representation structure of ViTs and CNNs on image classification benchmarks, we find striking differences between the two architectures, such as ViT having more uniform representations across all layers. We explore how these differences arise, finding crucial roles played by self-attention, which enables early aggregation of global information, and ViT residual connections, which strongly propagate features from lower to higher layers. We study the ramifications for spatial localization, demonstrating ViTs successfully preserve input spatial information, with noticeable effects from different classification methods. Finally, we study the effect of (pretraining) dataset scale on intermediate features and transfer learning, and conclude with a discussion on connections to new architectures such as the MLP-Mixer.

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。

↑