Exploring whether attention is necessary for vision transformers

Related tags

Deep Learningdo-you-even-need-attention
Overview

Do You Even Need Attention?
A Stack of Feed-Forward Layers Does Surprisingly Well on ImageNet

Paper/Report

TL;DR

We replace the attention layer in a vision transformer with a feed-forward layer and find that it still works quite well on ImageNet.

Abstract

The strong performance of vision transformers on image classification and other vision tasks is often attributed to the design of their multi-head attention layers. However, the extent to which attention is responsible for this strong performance remains unclear. In this short report, we ask: is the attention layer even necessary? Specifically, we replace the attention layer in a vision transformer with a feed-forward layer applied over the patch dimension. The resulting architecture is simply a series of feed-forward layers applied over the patch and feature dimensions in an alternating fashion. In experiments on ImageNet, this architecture performs surprisingly well: a ViT/DeiT-base-sized model obtains 74.9% top-1 accuracy, compared to 77.9% and 79.9% for ViT and DeiT respectively. These results indicate that aspects of vision transformers other than attention, such as the patch embedding, may be more responsible for their strong performance than previously thought. We hope these results prompt the community to spend more time trying to understand why our current models are as effective as they are.

Note

This is concurrent research with MLP-Mixer from Google Research. The ideas are exacty the same, with the one difference being that they use (a lot) more compute.

Pretrained models and logs

Here is a Weights and Biases report with the expected training trajectory: W&B

name [email protected] #params url
FF-tiny 61.4 7.7M model
FF-base 74.9 62M model
FF-large 71.4 206M -

Note: I haven't uploaded the FF-Large model because (1) it's over GitHub's file storage limit, and (2) I don't see why anyone would want it, given that it performs worse than the base model. That being said, if you want it, reach out to me and I'll send it to you.

How to train

The model definition in vision_transformer_linear.py is designed to be run with the repo from DeiT, which is itself based on the wonderful timm package.

Steps:

  • Clone the DeiT repo and move the file into it
git clone https://github.com/facebookresearch/deit
mv vision_transformer_linear.py deit
cd deit
  • Add a line to import vision_transformer_linear in main.py. For example, add the following after the import statements (around line 27):
+ import vision_transformer_linear
  • Train:
CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 python -m torch.distributed.launch \
--nproc_per_node=8 \
--master_port 10490 \
--use_env main.py \
--model linear_tiny \
--batch-size 128 \
--drop 0.1 \
--output_dir outputs/linear-tiny \
--data-path your/path/to/imagenet

Citation

If you build upon this idea, feel free to drop a citation (and also cite MLP-Mixer).

@article{melaskyriazi2021doyoueven,
  title={Do You Even Need Attention? A Stack of Feed-Forward Layers Does Surprisingly Well on ImageNet},
  author={Luke Melas-Kyriazi},
  journal=arxiv,
  year=2021
}
You might also like...
The first dataset of composite images with rationality score indicating whether the object placement in a composite image is reasonable.
The first dataset of composite images with rationality score indicating whether the object placement in a composite image is reasonable.

Object-Placement-Assessment-Dataset-OPA Object-Placement-Assessment (OPA) is to verify whether a composite image is plausible in terms of the object p

BCMI 53 Nov 15, 2022
Face Mask Detection is a project to determine whether someone is wearing mask or not, using deep neural network.
Face Mask Detection is a project to determine whether someone is wearing mask or not, using deep neural network.

face-mask-detection Face Mask Detection is a project to determine whether someone is wearing mask or not, using deep neural network. It contains 3 scr

amirsalar 13 Jan 18, 2022
This program was designed to detect whether someone is wearing a facemask through a live video stream.

This program was designed to detect whether someone is wearing a facemask through a live video stream. A custom lightweight CNN trained with TensorFlow on a public dataset provided by Kaggle is used to detect whether each face detected by the cv2 face detection dnn is wearing a mask

null 0 Apr 2, 2022
A system used to detect whether a person is wearing a medical mask or not.

Mask_Detection_System A system used to detect whether a person is wearing a medical mask or not. To open the program, please follow these steps: Make

Mohamed Emad 0 Nov 17, 2022
Diabet Feature Engineering - Predict whether people have diabetes when their characteristics are specified

Diabet Feature Engineering - Predict whether people have diabetes when their characteristics are specified

Şebnem 6 Jan 18, 2022
Diabetes-Feature-Engineering - A machine learning model that can predict whether people have diabetes when their characteristics are specified

Diabetes-Feature-Engineering Aim Developing a machine learning model that can pr

Özlem Taşkın 0 Feb 23, 2022
Multivariate Time Series Forecasting with efficient Transformers. Code for the paper
Multivariate Time Series Forecasting with efficient Transformers. Code for the paper "Long-Range Transformers for Dynamic Spatiotemporal Forecasting."

Spacetimeformer Multivariate Forecasting This repository contains the code for the paper, "Long-Range Transformers for Dynamic Spatiotemporal Forecast

QData 440 Jan 2, 2023
Implementation of SE3-Transformers for Equivariant Self-Attention, in Pytorch.

SE3 Transformer - Pytorch Implementation of SE3-Transformers for Equivariant Self-Attention, in Pytorch. May be needed for replicating Alphafold2 resu

Phil Wang 207 Dec 23, 2022
Official PyTorch implementation for Generic Attention-model Explainability for Interpreting Bi-Modal and Encoder-Decoder Transformers, a novel method to visualize any Transformer-based network. Including examples for DETR, VQA.
Official PyTorch implementation for Generic Attention-model Explainability for Interpreting Bi-Modal and Encoder-Decoder Transformers, a novel method to visualize any Transformer-based network. Including examples for DETR, VQA.

PyTorch Implementation of Generic Attention-model Explainability for Interpreting Bi-Modal and Encoder-Decoder Transformers 1 Using Colab Please notic

Hila Chefer 489 Jan 7, 2023
Comments
  • On details about the experiments

    On details about the experiments

    In the experiment section of the report, you mention:

    Such a comparison is not exactly fair because the feed-forward model uses stronger training augmentations.

    I wonder what the augmentations are, for details about the augmentation seems missing from the paper.

    opened by boredtylin 2
  • Positional Encoding Ablation

    Positional Encoding Ablation

    Hi Luke, thank you for sharing this amazing work.

    In your arxiv document, I cannot find any mention of positional encoding, but I see that you use them in your code. Did you conduct any ablation study on the PE? i.e., how much does it affect the performance, with and without?

    Thank you in advance.

    opened by jjparkcv 0
  • Interaction between patches through a transpose may have a stronger role to play ?

    Interaction between patches through a transpose may have a stronger role to play ?

    Hi, I was going through your exp report. You have made a point that since you are able to get a good performance without using attention layer so good performance of ViT may be more to do with it's embedding layer than attention .

    But I believe, It's also may be to do with how you have established an interaction between patches through a transpose very similar to what was done in MLP-Mixer .

    Would love to know your thoughts on this ?

    opened by rakshith291 0
Owner
Luke Melas-Kyriazi
I'm student at Harvard University studying mathematics and computer science, always open to collaborate on interesting projects!
Luke Melas-Kyriazi
GitHub Repository
SASM - simple crossplatform IDE for NASM, MASM, GAS and FASM assembly languages

SASM (SimpleASM) - простая кроссплатформенная среда разработки для языков ассемблера NASM, MASM, GAS, FASM с подсветкой синтаксиса и отладчиком. В SA

Dmitriy Manushin 5.6k Jan 06, 2023
An image base contains 490 images for learning (400 cars and 90 boats), and another 21 images for testingAn image base contains 490 images for learning (400 cars and 90 boats), and another 21 images for testing

SVM Données Une base d’images contient 490 images pour l’apprentissage (400 voitures et 90 bateaux), et encore 21 images pour fait des tests. Prétrait

Achraf Rahouti 3 Nov 30, 2021
"Structure-Augmented Text Representation Learning for Efficient Knowledge Graph Completion"(WWW 2021)

STAR_KGC This repo contains the source code of the paper accepted by WWW'2021. "Structure-Augmented Text Representation Learning for Efficient Knowled

Bo Wang 60 Dec 26, 2022
Streamlit component for TensorBoard, TensorFlow's visualization toolkit

streamlit-tensorboard This is a work-in-progress, providing a function to embed TensorBoard, TensorFlow's visualization toolkit, in Streamlit apps. In

Snehan Kekre 27 Nov 13, 2022
Flower - A Friendly Federated Learning Framework

Flower - A Friendly Federated Learning Framework Flower (flwr) is a framework for building federated learning systems. The design of Flower is based o

Adap 1.8k Jan 01, 2023
Face Alignment using python

Face Alignment Face Alignment using python Input Image Aligned Face Aligned Face Aligned Face Input Image Aligned Face Input Image Aligned Face Instal

Sajjad Aemmi 28 Nov 23, 2022
[TNNLS 2021] The official code for the paper "Learning Deep Context-Sensitive Decomposition for Low-Light Image Enhancement"

CSDNet-CSDGAN this is the code for the paper "Learning Deep Context-Sensitive Decomposition for Low-Light Image Enhancement" Environment Preparing pyt

Jiaao Zhang 17 Nov 05, 2022
Language-Agnostic Website Embedding and Classification

Homepage2Vec Language-Agnostic Website Embedding and Classification based on Curlie labels https://arxiv.org/pdf/2201.03677.pdf Homepage2Vec is a pre-

25 Dec 27, 2022
NeRF visualization library under construction

NeRF visualization library using PlenOctrees, under construction pip install nerfvis Docs will be at: https://nerfvis.readthedocs.org import nerfvis s

Alex Yu 196 Jan 04, 2023
Python implementation of ADD: Frequency Attention and Multi-View based Knowledge Distillation to Detect Low-Quality Compressed Deepfake Images, AAAI2022.

ADD: Frequency Attention and Multi-View based Knowledge Distillation to Detect Low-Quality Compressed Deepfake Images Binh M. Le & Simon S. Woo, "ADD:

2 Oct 24, 2022
Marine debris detection with commercial satellite imagery and deep learning.

Marine debris detection with commercial satellite imagery and deep learning. Floating marine debris is a global pollution problem which threatens mari

Inter Agency Implementation and Advanced Concepts 56 Dec 16, 2022
Mosaic of Object-centric Images as Scene-centric Images (MosaicOS) for long-tailed object detection and instance segmentation.

MosaicOS Mosaic of Object-centric Images as Scene-centric Images (MosaicOS) for long-tailed object detection and instance segmentation. Introduction M

Cheng Zhang 27 Oct 12, 2022
Finetuner allows one to tune the weights of any deep neural network for better embeddings on search tasks

Finetuner allows one to tune the weights of any deep neural network for better embeddings on search tasks

Jina AI 794 Dec 31, 2022
A tf.keras implementation of Facebook AI's MadGrad optimization algorithm

MADGRAD Optimization Algorithm For Tensorflow This package implements the MadGrad Algorithm proposed in Adaptivity without Compromise: A Momentumized,

20 Aug 18, 2022
The project of phase's key role in complex and real NN

Phase-in-NN This is the code for our project at Princeton (co-authors: Yuqi Nie, Hui Yuan). The paper title is: "Neural Network is heterogeneous: Phas

YuqiNie-lab 1 Nov 04, 2021
Unofficial PyTorch implementation of Fastformer based on paper "Fastformer: Additive Attention Can Be All You Need"."

Fastformer-PyTorch Unofficial PyTorch implementation of Fastformer based on paper Fastformer: Additive Attention Can Be All You Need. Usage : import t

Hong-Jia Chen 126 Dec 06, 2022
Self-Guided Contrastive Learning for BERT Sentence Representations

Self-Guided Contrastive Learning for BERT Sentence Representations This repository is dedicated for releasing the implementation of the models utilize

Taeuk Kim 16 Dec 04, 2022
This is the code repository for the paper A hierarchical semantic segmentation framework for computer-vision-based bridge column damage detection

Bridge-damage-segmentation This is the code repository for the paper A hierarchical semantic segmentation framework for computer-vision-based bridge c

Jingxiao Liu 5 Dec 07, 2022
TrTr: Visual Tracking with Transformer

TrTr: Visual Tracking with Transformer We propose a novel tracker network based on a powerful attention mechanism called Transformer encoder-decoder a

趙　漠居(Zhao, Moju) 66 Dec 27, 2022
Official pytorch implementation of "DSPoint: Dual-scale Point Cloud Recognition with High-frequency Fusion"

DSPoint Official implementation of "DSPoint: Dual-scale Point Cloud Recognition with High-frequency Fusion". Paper link: https://arxiv.org/abs/2111.10

Ziyao Zeng 14 Feb 26, 2022