The official implementation of ELSA: Enhanced Local Self-Attention for Vision Transformer

Related tags

Deep LearningELSA
Overview

ELSA: Enhanced Local Self-Attention for Vision Transformer

By Jingkai Zhou, Pichao Wang*, Fan Wang, Qiong Liu, Hao Li, Rong Jin

This repo is the official implementation of "ELSA: Enhanced Local Self-Attention for Vision Transformer".

Introduction

Self-attention is powerful in modeling long-range dependencies, but it is weak in local finer-level feature learning. As shown in Figure 1, the performance of local self-attention (LSA) is just on par with convolution and inferior to dynamic filters, which puzzles researchers on whether to use LSA or its counterparts, which one is better, and what makes LSA mediocre. In this work, we comprehensively investigate LSA and its counterparts. We find that the devil lies in the generation and application of spatial attention.

Based on these findings, we propose the enhanced local self-attention (ELSA) with Hadamard attention and the ghost head, as illustrated in Figure 2. Experiments demonstrate the effectiveness of ELSA. Without architecture / hyperparameter modification, The use of ELSA in drop-in replacement boosts baseline methods consistently in both upstream and downstream tasks.

Please refer to our paper for more details.

Model zoo

ImageNet Classification

Model #Params Pretrain Resolution Top1 Acc Download
ELSA-Swin-T 28M ImageNet 1K 224 82.7 google / baidu
ELSA-Swin-S 53M ImageNet 1K 224 83.5 google / baidu
ELSA-Swin-B 93M ImageNet 1K 224 84.0 google / baidu

COCO Object Detection

Backbone Method Pretrain Lr Schd Box mAP Mask mAP #Params Download
ELSA-Swin-T Mask R-CNN ImageNet-1K 1x 45.7 41.1 49M google / baidu
ELSA-Swin-T Mask R-CNN ImageNet-1K 3x 47.5 42.7 49M google / baidu
ELSA-Swin-S Mask R-CNN ImageNet-1K 1x 48.3 43.0 72M google / baidu
ELSA-Swin-S Mask R-CNN ImageNet-1K 3x 49.2 43.6 72M google / baidu
ELSA-Swin-T Cascade Mask R-CNN ImageNet-1K 1x 49.8 43.0 86M google / baidu
ELSA-Swin-T Cascade Mask R-CNN ImageNet-1K 3x 51.0 44.2 86M google / baidu
ELSA-Swin-S Cascade Mask R-CNN ImageNet-1K 1x 51.6 44.4 110M google / baidu
ELSA-Swin-S Cascade Mask R-CNN ImageNet-1K 3x 52.3 45.2 110M google / baidu

ADE20K Semantic Segmentation

Backbone Method Pretrain Crop Size Lr Schd mIoU (ms+flip) #Params Download
ELSA-Swin-T UPerNet ImageNet-1K 512x512 160K 47.9 61M google / baidu
ELSA-Swin-S UperNet ImageNet-1K 512x512 160K 50.4 85M google / baidu

Install

  • Clone this repo:
git clone https://github.com/damo-cv/ELSA.git elsa
cd elsa
  • Create a conda virtual environment and activate it:
conda create -n elsa python=3.7 -y
conda activate elsa
  • Install PyTorch==1.8.0 and torchvision==0.9.0 with CUDA==10.1:
conda install pytorch==1.8.0 torchvision==0.9.0 cudatoolkit=10.1 -c pytorch
git clone https://github.com/NVIDIA/apex
cd apex
pip install -v --disable-pip-version-check --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./
cd ../
  • Install mmcv-full==1.3.0
pip install mmcv-full==1.3.0 -f https://download.openmmlab.com/mmcv/dist/cu101/torch1.8.0/index.html
  • Install other requirements:
pip install -r requirements.txt
  • Install mmdet and mmseg:
cd ./det
pip install -v -e .
cd ../seg
pip install -v -e .
cd ../
  • Build the elsa operation:
cd ./cls/models/elsa
python setup.py install
mv build/lib*/* .
cp *.so ../../../det/mmdet/models/backbones/elsa/
cp *.so ../../../seg/mmseg/models/backbones/elsa/
cd ../../../

Data preparation

We use standard ImageNet dataset, you can download it from http://image-net.org/. Please prepare it under the following file structure:

$ tree data
imagenet
├── train
│   ├── class1
│   │   ├── img1.jpeg
│   │   ├── img2.jpeg
│   │   └── ...
│   ├── class2
│   │   ├── img3.jpeg
│   │   └── ...
│   └── ...
└── val
    ├── class1
    │   ├── img4.jpeg
    │   ├── img5.jpeg
    │   └── ...
    ├── class2
    │   ├── img6.jpeg
    │   └── ...
    └── ...

Also, please prepare the COCO and ADE20K datasets following their links. Then, please link them to det/data and seg/data.

Evaluation

ImageNet Classification

Run following scripts to evaluate pre-trained models on the ImageNet-1K:

cd cls

python validate.py <PATH_TO_IMAGENET> --model elsa_swin_tiny --checkpoint <CHECKPOINT_FILE> \
  --no-test-pool --apex-amp --img-size 224 -b 128

python validate.py <PATH_TO_IMAGENET> --model elsa_swin_small --checkpoint <CHECKPOINT_FILE> \
  --no-test-pool --apex-amp --img-size 224 -b 128

python validate.py <PATH_TO_IMAGENET> --model elsa_swin_base --checkpoint <CHECKPOINT_FILE> \
  --no-test-pool --apex-amp --img-size 224 -b 128 --use-ema

COCO Detection

Run following scripts to evaluate a detector on the COCO:

cd det

# single-gpu testing
python tools/test.py <CONFIG_FILE> <DET_CHECKPOINT_FILE> --eval bbox segm

# multi-gpu testing
tools/dist_test.sh <CONFIG_FILE> <DET_CHECKPOINT_FILE> <GPU_NUM> --eval bbox segm

ADE20K Semantic Segmentation

Run following scripts to evaluate a model on the ADE20K:

cd seg

# single-gpu testing
python tools/test.py <CONFIG_FILE> <SEG_CHECKPOINT_FILE> --aug-test --eval mIoU

# multi-gpu testing
tools/dist_test.sh <CONFIG_FILE> <SEG_CHECKPOINT_FILE> <GPU_NUM> --aug-test --eval mIoU

Training from scratch

Due to randomness, the re-training results may have a gap of about 0.1~0.2% with the numbers in the paper.

ImageNet Classification

Run following scripts to train classifiers on the ImageNet-1K:

cd cls

bash ./distributed_train.sh 8 <PATH_TO_IMAGENET> --model elsa_swin_tiny \
  --epochs 300 -b 128 -j 8 --opt adamw --lr 1e-3 --sched cosine --weight-decay 5e-2 \
  --warmup-epochs 20 --warmup-lr 1e-6 --min-lr 1e-5 --drop-path 0.1 --aa rand-m9-mstd0.5-inc1 \
  --mixup 0.8 --cutmix 1. --remode pixel --reprob 0.25 --clip-grad 5. --amp

bash ./distributed_train.sh 8 <PATH_TO_IMAGENET> --model elsa_swin_small \
  --epochs 300 -b 128 -j 8 --opt adamw --lr 1e-3 --sched cosine --weight-decay 5e-2 \
  --warmup-epochs 20 --warmup-lr 1e-6 --min-lr 1e-5 --drop-path 0.3 --aa rand-m9-mstd0.5-inc1 \
  --mixup 0.8 --cutmix 1. --remode pixel --reprob 0.25 --clip-grad 5. --amp

bash ./distributed_train.sh 8 <PATH_TO_IMAGENET> --model elsa_swin_base \
  --epochs 300 -b 128 -j 8 --opt adamw --lr 1e-3 --sched cosine --weight-decay 5e-2 \
  --warmup-epochs 20 --warmup-lr 1e-6 --min-lr 1e-5 --drop-path 0.5 --aa rand-m9-mstd0.5-inc1 \
  --mixup 0.8 --cutmix 1. --remode pixel --reprob 0.25 --clip-grad 5. --amp --model-ema

If GPU memory is not enough when training elsa_swin_base, you can use two nodes (2 * 8 GPUs), each with a batch size of 64 images/GPU.

COCO Detection / ADE20K Semantic Segmentation

Run following scripts to train models on the COCO / ADE20K:

cd det 
# (or cd seg)

# multi-gpu training
tools/dist_train.sh <CONFIG_FILE> <GPU_NUM> --cfg-options model.pretrained=<PRETRAIN_MODEL> [model.backbone.use_checkpoint=True] [other optional arguments]

Acknowledgement

This work was supported by Alibaba Group through Alibaba Research Intern Program and the National Natural Science Foundation of China (No.61976094).

Codebase from pytorch-image-models, ddfnet, VOLO, Swin-Transformer, Swin-Transformer-Detection, and Swin-Transformer-Semantic-Segmentation

Citing ELSA

@article{zhou2021ELSA,
  title={ELSA: Enhanced Local Self-Attention for Vision Transformer},
  author={Zhou, Jingkai and Wang, Pichao and Wang, Fan and Liu, Qiong and Li, Hao and Jin, Rong},
  journal={arXiv preprint arXiv:2112.12786},
  year={2021}
}
Owner
DamoCV
CV team of DAMO academy
DamoCV
GitHub Repository
A task Provided by A respective Artenal Ai and Ml based Company to complete it

A task Provided by A respective Alternal Ai and Ml based Company to complete it .

Parth Madan 1 Jan 25, 2022
[AAAI-2022] Official implementations of MCL: Mutual Contrastive Learning for Visual Representation Learning

Mutual Contrastive Learning for Visual Representation Learning This project provides source code for our Mutual Contrastive Learning for Visual Repres

winycg 48 Jan 02, 2023
Pywonderland - A tour in the wonderland of math with python.

A Tour in the Wonderland of Math with Python A collection of python scripts for drawing beautiful figures and animating interesting algorithms in math

Zhao Liang 4.1k Jan 03, 2023
This is the repo for Uncertainty Quantification 360 Toolkit.

UQ360 The Uncertainty Quantification 360 (UQ360) toolkit is an open-source Python package that provides a diverse set of algorithms to quantify uncert

International Business Machines 207 Dec 30, 2022
Theano is a Python library that allows you to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficiently. It can use GPUs and perform efficient symbolic differentiation.

============================================================================================================ `MILA will stop developing Theano https:

9.6k Dec 31, 2022
Official code for our ICCV paper: "From Continuity to Editability: Inverting GANs with Consecutive Images"

GANInversion_with_ConsecutiveImgs Official code for our ICCV paper: "From Continuity to Editability: Inverting GANs with Consecutive Images" https://a

QingyangXu 38 Dec 07, 2022
Human Activity Recognition example using TensorFlow on smartphone sensors dataset and an LSTM RNN. Classifying the type of movement amongst six activity categories - Guillaume Chevalier

LSTMs for Human Activity Recognition Human Activity Recognition (HAR) using smartphones dataset and an LSTM RNN. Classifying the type of movement amon

Guillaume Chevalier 3.1k Dec 30, 2022
Semi-Supervised Signed Clustering Graph Neural Network (and Implementation of Some Spectral Methods)

SSSNET SSSNET: Semi-Supervised Signed Network Clustering For details, please read our paper. Environment Setup Overview The project has been tested on

Yixuan He 9 Nov 24, 2022
Implementation of the ICCV'21 paper Temporally-Coherent Surface Reconstruction via Metric-Consistent Atlases

Temporally-Coherent Surface Reconstruction via Metric-Consistent Atlases [Papers 1, 2][Project page] [Video] The implementation of the papers Temporal

56 Nov 21, 2022
Angular & Electron desktop UI framework. Angular components for native looking and behaving macOS desktop UI (Electron/Web)

Angular Desktop UI This is a collection for native desktop like user interface components in Angular, especially useful for Electron apps. It starts w

Marc J. Schmidt 49 Dec 22, 2022
Image Restoration Using Swin Transformer for VapourSynth

SwinIR SwinIR function for VapourSynth, based on https://github.com/JingyunLiang/SwinIR. Dependencies NumPy PyTorch, preferably with CUDA. Note that t

Holy Wu 11 Jun 19, 2022
Out-of-boundary View Synthesis towards Full-frame Video Stabilization

Out-of-boundary View Synthesis towards Full-frame Video Stabilization Introduction | Update | Results Demo | Introduction This repository contains the

25 Oct 10, 2022
Official implementation of Representer Point Selection via Local Jacobian Expansion for Post-hoc Classifier Explanation of Deep Neural Networks and Ensemble Models at NeurIPS 2021

Representer Point Selection via Local Jacobian Expansion for Classifier Explanation of Deep Neural Networks and Ensemble Models This repository is the

Yi(Amy) Sui 2 Dec 01, 2021
Code for Ditto: Building Digital Twins of Articulated Objects from Interaction

Ditto: Building Digital Twins of Articulated Objects from Interaction Zhenyu Jiang, Cheng-Chun Hsu, Yuke Zhu CVPR 2022, Oral Project | arxiv News 2022

UT Robot Perception and Learning Lab 78 Dec 22, 2022
Source Code for AAAI 2022 paper "Graph Convolutional Networks with Dual Message Passing for Subgraph Isomorphism Counting and Matching"

Graph Convolutional Networks with Dual Message Passing for Subgraph Isomorphism Counting and Matching This repository is an official implementation of

HKUST-KnowComp 13 Sep 08, 2022
Graph parsing approach to structured sentiment analysis.

Fine-grained Sentiment Analysis as Dependency Graph Parsing This repository contains the code and datasets described in following paper: Fine-grained

Jeremy Barnes 36 Dec 12, 2022
Working demo of the Multi-class and Anomaly classification model using the CLIP feature space

👁️ Hindsight AI: Crime Classification With Clip About For Educational Purposes Only This is a recursive neural net trained to classify specific crime

Miles Tweed 2 Jun 05, 2022
Using contrastive learning and OpenAI's CLIP to find good embeddings for images with lossy transformations

The official code for the paper "Inverse Problems Leveraging Pre-trained Contrastive Representations" (to appear in NeurIPS 2021).

Sriram Ravula 26 Dec 10, 2022
Implementation of Self-supervised Graph-level Representation Learning with Local and Global Structure (ICML 2021).

Self-supervised Graph-level Representation Learning with Local and Global Structure Introduction This project is an implementation of ``Self-supervise

MilaGraph 50 Dec 09, 2022
Implementation of the paper: "SinGAN: Learning a Generative Model from a Single Natural Image"

SinGAN This is an unofficial implementation of SinGAN from someone who's been sitting right next to SinGAN's creator for almost five years. Please ref

35 Nov 10, 2022