Do Vision Transformers See Like Convolutional Neural Networks?

08/19/2021
by   Jojo Yun, et al.
33

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.

READ FULL TEXT

page 3

page 7

page 8

page 16

page 23

page 24

page 25

page 29

research
12/20/2014

Visualizing and Comparing Convolutional Neural Networks

Convolutional Neural Networks (CNNs) have achieved comparable error rate...
research
12/13/2022

What do Vision Transformers Learn? A Visual Exploration

Vision transformers (ViTs) are quickly becoming the de-facto architectur...
research
10/13/2022

Vision Transformers provably learn spatial structure

Vision Transformers (ViTs) have achieved comparable or superior performa...
research
03/01/2022

Tricks and Plugins to GBM on Images and Sequences

Convolutional neural networks (CNNs) and transformers, which are compose...
research
12/09/2022

Mitigation of Spatial Nonstationarity with Vision Transformers

Spatial nonstationarity, the location variance of features' statistical ...
research
11/26/2022

Towards Better Input Masking for Convolutional Neural Networks

The ability to remove features from the input of machine learning models...
research
09/19/2023

Exploring the Influence of Information Entropy Change in Learning Systems

In this work, we explore the influence of entropy change in deep learnin...

Please sign up or login with your details

Forgot password? Click here to reset