neverstoplearn/FOTS.PyTorch

FOTS Pytorch Implementation

News!!! Recognition branch now is added into model. The whole project has beed optimized and refactored.

• ICDAR Dataset
• SynthText 800K Dataset
• detection branch (verified on the training set, It works!)
• recognition branch (verified)
• eval
• multi-gpu training
• reasonable project structure
• wandb
• pytorch_lightning

Introduction

This is a PyTorch implementation of FOTS.

Instruction

Requirements

1. build tools

   ./build.sh
   

2. prepare Dataset

3. create virtual env, you may need conda

   conda create --name fots --file spec-file.txt
   conda activate fots
   pip install -r reqs.txt
   

Training

# quite easy, for single gpu training set gpus to [0]. 0 is the id of your gpu.
python train.py -c pretrain.json
python train.py -c finetune.json

Evaluation

python eval.py -m <model.tar.gz> -i <input_images_folder> -o <output_folders>

Benchmarking and Models (Coming soon!)

Visualization (1000 epochs, 8 bs, icdar2015 without finetuning, still converging!!!)

Acknowledgement

• https://github.com/SakuraRiven/EAST (Some codes are copied from here.)
• https://github.com/chenjun2hao/FOTS.pytorch.git (ROIRotate)