Improving Transferability of Representations via Augmentation-Aware Self-Supervision

Last update: Sep 16, 2022

Related tags

Overview

Improving Transferability of Representations via Augmentation-Aware Self-Supervision

Accepted to NeurIPS 2021

TL;DR: Learning augmentation-aware information by predicting the difference between two augmented samples improves the transferability of representations.

Dependencies

conda create -n AugSelf python=3.8 pytorch=1.7.1 torchvision=0.8.2 cudatoolkit=10.1 ignite -c pytorch
conda activate AugSelf
pip install scipy tensorboard kornia==0.4.1 sklearn

Checkpoints

We provide ImageNet100-pretrained models in this Dropbox link.

Pretraining

We here provide SimSiam+AugSelf pretraining scripts. For training the baseline (i.e., no AugSelf), remove --ss-crop and --ss-color options. For using other frameworks like SimCLR, use the --framework option.

STL-10

CUDA_VISIBLE_DEVICES=0 python pretrain.py \
    --logdir ./logs/stl10/simsiam/aug_self \
    --framework simsiam \
    --dataset stl10 \
    --datadir DATADIR \
    --model resnet18 \
    --batch-size 256 \
    --max-epochs 200 \
    --ss-color 1.0 --ss-crop 1.0

ImageNet100

python pretrain.py \
    --logdir ./logs/imagenet100/simsiam/aug_self \
    --framework simsiam \
    --dataset imagenet100 \
    --datadir DATADIR \
    --batch-size 256 \
    --max-epochs 500 \
    --model resnet50 \
    --base-lr 0.05 --wd 1e-4 \
    --ckpt-freq 50 --eval-freq 50 \
    --ss-crop 0.5 --ss-color 0.5 \
    --num-workers 16 --distributed

Evaluation

Our main evaluation setups are linear evaluation on fine-grained classification datasets (Table 1) and few-shot benchmarks (Table 2).

linear evaluation

CUDA_VISIBLE_DEVICES=0 python transfer_linear_eval.py \
    --pretrain-data imagenet100 \
    --ckpt CKPT \
    --model resnet50 \
    --dataset cifar10 \
    --datadir DATADIR \
    --metric top1

few-shot

CUDA_VISIBLE_DEVICES=0 python transfer_few_shot.py \
    --pretrain-data imagenet100 \
    --ckpt CKPT \
    --model resnet50 \
    --dataset cub200 \
    --datadir DATADIR

Improving Transferability of Representations via Augmentation-Aware Self-Supervision

Related tags

Overview

Improving Transferability of Representations via Augmentation-Aware Self-Supervision

Dependencies

Checkpoints

Pretraining

STL-10

ImageNet100

Evaluation

linear evaluation

few-shot

Owner

hankook

A clean and scalable template to kickstart your deep learning project 🚀 ⚡ 🔥

Code for "Multi-View Multi-Person 3D Pose Estimation with Plane Sweep Stereo"

Robot Reinforcement Learning on the Constraint Manifold

Official implementation of "Dynamic Anchor Learning for Arbitrary-Oriented Object Detection" (AAAI2021).

Automatic 2D-to-3D Video Conversion with CNNs

Devkit for 3D -- Some utils for 3D object detection based on Numpy and Pytorch

A python program to hack instagram

CVPR 2021 - Official code repository for the paper: On Self-Contact and Human Pose.

Code for our TKDE paper "Understanding WeChat User Preferences and “Wow” Diffusion"

Yas CRNN model training - Yet Another Genshin Impact Scanner

Neural Factorization of Shape and Reflectance Under An Unknown Illumination

A collection of Jupyter notebooks to play with NVIDIA's StyleGAN3 and OpenAI's CLIP for a text-based guided image generation.

Simulation of self-focusing of laser beams in condensed media

This repository contains an overview of important follow-up works based on the original Vision Transformer (ViT) by Google.

Code for models used in Bashiri et al., "A Flow-based latent state generative model of neural population responses to natural images".

TorchPQ is a python library for Approximate Nearest Neighbor Search (ANNS) and Maximum Inner Product Search (MIPS) on GPU using Product Quantization (PQ) algorithm.

WormMovementSimulation - 3D Simulation of Worm Body Movement with Neurons attached to its body

An official source code for "Augmentation-Free Self-Supervised Learning on Graphs"

Differentiable Prompt Makes Pre-trained Language Models Better Few-shot Learners

Code accompanying the paper "Knowledge Base Completion Meets Transfer Learning"