https://github.com/lromul/argus-alaska

Kaggle | Part of 25th place solution for ALASKA2 Image Steganalysis kaggle competition.

Science Score: 13.0%

This score indicates how likely this project is to be science-related based on various indicators:

○
CITATION.cff file
✓
codemeta.json file
Found codemeta.json file
○
.zenodo.json file
○
DOI references
○
Academic publication links
○
Committers with academic emails
○
Institutional organization owner
○
JOSS paper metadata
○
Scientific vocabulary similarity
Low similarity (6.3%) to scientific vocabulary

Keywords

computer-vision deep-learning kaggle kaggle-competition pytorch steganalysis

Keywords from Contributors

archival projection profiles optim interactive sequences generic autograding hacking shellcodes

Last synced: 6 months ago · JSON representation

Repository

Kaggle | Part of 25th place solution for ALASKA2 Image Steganalysis kaggle competition.

Basic Info

Host: GitHub
Owner: lRomul
License: mit
Language: Python
Default Branch: master
Homepage: https://www.kaggle.com/c/alaska2-image-steganalysis
Size: 94.7 KB

Statistics

Stars: 3
Watchers: 2
Forks: 0
Open Issues: 4
Releases: 0

Topics

computer-vision deep-learning kaggle kaggle-competition pytorch steganalysis

Created almost 6 years ago · Last pushed almost 3 years ago

Metadata Files

Readme License

README.md

ALASKA2 Image Steganalysis

Source code of solution for ALASKA2 Image Steganalysis competition.

Solution

Key points: * Efficientnets * DDP training with SyncBN and Apex mixed precision * AdamW with cosine annealing * EMA Model * Bitmix

Quick setup and start

Requirements

Nvidia drivers, CUDA >= 10.2, cuDNN >= 7
Docker, nvidia-docker

The provided dockerfile is supplied to build image with cuda support and cudnn.

Preparations

Clone the repo, build docker image. bash git clone https://github.com/lRomul/argus-alaska.git cd argus-alaska make build
Download and extract dataset to data folder.

Run

Run docker container bash make
Create folds split and extract quality of images bash python make_folds.py python make_quality_json.py
Train model bash python train.py --experiment train_001
Predict test and make submission bash python predict.py --experiment train_001
Train on 4 GPUs with distributed data parallel bash ./distributed_train.sh 4 --experiment ddp_train_001

Owner

Name: Ruslan Baikulov
Login: lRomul
Kind: user
Location: Moscow, Russia

Repositories: 5
Profile: https://github.com/lRomul

Deep Learning Engineer

GitHub Events

Total

Last Year

Committers

Last synced: 10 months ago

All Time

Total Commits: 124
Total Committers: 2
Avg Commits per committer: 62.0
Development Distribution Score (DDS): 0.008

Past Year

Commits: 0
Committers: 0
Avg Commits per committer: 0.0
Development Distribution Score (DDS): 0.0

Top Committers

Name	Email	Commits
ruslan	r**3@g**m	123
dependabot[bot]	4****]	1

Issues and Pull Requests

Last synced: about 1 year ago

All Time

Total issues: 0
Total pull requests: 7
Average time to close issues: N/A
Average time to close pull requests: 5 months
Total issue authors: 0
Total pull request authors: 1
Average comments per issue: 0
Average comments per pull request: 0.29
Merged pull requests: 1
Bot issues: 0
Bot pull requests: 7

Past Year

Issues: 0
Pull requests: 0
Average time to close issues: N/A
Average time to close pull requests: N/A
Issue authors: 0
Pull request authors: 0
Average comments per issue: 0
Average comments per pull request: 0
Merged pull requests: 0
Bot issues: 0
Bot pull requests: 0

View more stats

Top Authors

Issue Authors

Pull Request Authors

dependabot[bot] (7)

Top Labels

Issue Labels

Pull Request Labels

dependencies (7)

Dependencies

requirements.txt pypi

Cython ==0.29.17
adamp ==0.2.0
albumentations ==0.4.5
matplotlib ==3.2.1
notebook ==6.1.5
numpy ==1.18.4
opencv-python ==4.2.0.34
pandas ==1.0.3
pytorch-argus ==0.1.2
scikit-image ==0.16.2
scikit-learn ==0.22.2.post1
scipy ==1.4.1
timm ==0.1.26
torch ==1.5.1
torchvision ==0.6.0

Dockerfile docker

nvidia/cuda 10.2-cudnn7-devel-ubuntu18.04 build