bcaitech1 / p4-fr-soccer

p4-fr-soccer created by GitHub Classroom
1 stars 3 forks source link

Team sOcCeR's formular recognizer(OCR)

목차

프로젝트 기간

프로젝트 overview

All of what we experiments

last structure

Flow of what we applied

image

More Details

자세한 실험과정과 추가적인 내용들은 Project Details를 확인해주세요

수식 인식기 프로젝트 소개

formular

평가 방법

Environment

Hardware

Software

Dependencies

$ pip install -r requirements.txt

Usage

Attention, SATRN

$ python train.py --c {your_model}.yaml

Vit

$ python train_ViT.py

Swin

$ python train_swin.py

Dataset

학습이미지 예시:\ image

Ground Truth:\ x = \frac { - b \pm \sqrt { b ^ 2 - 4 a c } } { 2 a } \ { \text { when } } \ {a x ^ 2 + b x + c = 0}

File Structure

p4-fr-hatting-day/code/
│
├── configs
│    ├── Attention.yaml
│    ├── ...
│    └── EFFICIENT_SATRNv6.yaml
├── datatools
│    ├── extract_tokens.py
│    ├── parse_upstage.py
│    └── train_test_split.py
├── network
│    ├── Attention.py
│    ├── EFFICIENT_SATRN.py
│    ├── SATRN_extension.py
│    ├── ...
│    └── swin.py
├── submit
├── checkpoint.py
├── dataset.py
├── dataset_ViT.py
├── dataset_Swin.py
├── floags.py
├── inference.py
├── inference_ensemble.py
├── metrics.py
├── requirements.txt
├── requirements_2.txt
├── scheduler.py
├── submission.txt
├── train.py
├── train_ViT.py
├── train_swin.py
└── utils.py
p4-fr-hatting-day/pytorch-CycleGAN-and-pix2pix/
├── CycleGAN.ipynb # for tutorial
├── ...
└── cleaning_GAN.ipynb # for data cleansing

Models

ASTER

SATRN

ViT

Swin

Data cleaning with GAN

How to use

Sample Results

Contributors

Reference

Papers

Supported Data