xiaoleidiao / lowlevelvision-pipeline-pytorch

An easy-handle pipeline project for low level vision tasks (Image2Image) by pytorch

LowLevelVision-Pipeline-pytorch

This is an easy-handle pipeline project for common AI tasks by pytorch. Currently, this project is mainly about low level vision tasks (Image2Image)

Files:

1. datasets.py: loading images from target path
2. train.py: Model training pipeline
3. predict.py: task processing pipeline

Folders:

1. data: datasets for training and testing
2. models: different NN models
3. saved_models: trained models weights
4. validImages: validition of test images for visualization
5. util: Tools for utilizing the model training, or additional processing like visualization.
6. wandb: wandb recoder for result/parameter visualization (optional)

Requirements:

• CUDA==10.2
• cudnn==8.0
• python==3.6
• torch==1.8.0
• torchvision==0.9.0
• DNN-printer==0.0.2
• numpy
• tqdm
• wandb (optional)
• argparse
• wmi (optional)

################# test