/fastdvdnet

FastDVDnet + Docker + Jupyter Notebook + Colab

Primary LanguagePython

FastDVDnet + Docker

# Clone the repo
git clone https://github.com/ZurMaD/fastdvdnet

# Change directory
cd fastdvdnet/

# Download docker with requirements
docker pull pablogod/fastdvdnet

# Remove all images inside path
rm ./upload_images/*

# Split the video on frames like below 'upload_images' is where the frames are being update
# You need to specify framerate on '-r 1/1'
ffmpeg -i path/to/your_video/1.mp4 -r 1/1 /content/fastdvdnet/upload_images/%03d.png

# Check https://github.com/anibali/docker-pytorch for full information
docker run --rm -it --init \
  --gpus=all \
  --ipc=host \
  --user="$(id -u):$(id -g)" \
  --volume="$PWD:/app" \ # NOW EXECUTE THE PY FILE WITH ARGS
  pablogod/fastdvdnet python3 test_fastdvdnet.py \ 
	--test_path ./upload_images \ # path inside git cloned. Need a video splitted on frames with numbers like 001.png, 002.png
	--noise_sigma 30 \ # [5,45] Check paper to understand this part
	--save_path ./results # Folder that the results are going to be storage
	
#Notes:
#   - Code is not working with CPU only
#   - Needs NVIDIA drive support before run the docker correctly

FastDVDnet

A state-of-the-art, simple and fast network for Deep Video Denoising which uses no motion compensation.

NEW: Paper to be presented at CVPR2020

Previous deep video denoising algorithm: DVDnet

Overview

This source code provides a PyTorch implementation of the FastDVDnet video denoising algorithm, as in Tassano, Matias and Delon, Julie and Veit, Thomas. "FastDVDnet: Towards Real-Time Deep Video Denoising Without Flow Estimation", arXiv preprint arXiv:1907.01361 (2019).

Video Examples

You can download several denoised sequences with our algorithm and other methods here (more videos coming soon)

Running Times

FastDVDnet is orders of magnitude faster than other state-of-the-art methods

Results

Left: Input noise sigma 40 denoised with FastDVDnet (sorry about the dithering due to gif compression)

Right: PSNRs on the DAVIS testset, Gaussian noise and clipped Gaussian noise

Architecture

Code User Guide

Dependencies

The code runs on Python +3.6. You can create a conda environment with all the dependecies by running (Thanks to Antoine Monod for the .yml file)

conda env create -f requirements.yml -n <env_name>

Note: this project needs the NVIDIA DALI package for training. The tested version of DALI is 0.10.0. If you prefer to install it yourself (supposing you have CUDA 10.0), you need to run

pip install --extra-index-url https://developer.download.nvidia.com/compute/redist/cuda/10.0 nvidia-dali==0.10.0 

Testing

If you want to denoise an image sequence using the pretrained model you can execute

test_fastdvdnet.py \
	--test_path <path_to_input_sequence> \
	--noise_sigma 30 \
	--save_path results

NOTES

  • The image sequence should be stored under <path_to_input_sequence>
  • The model has been trained for values of noise in [5, 55]
  • run with --no_gpu to run on CPU instead of GPU
  • run with --save_noisy to save noisy frames
  • set max_num_fr_per_seq to set the max number of frames to load per sequence
  • to denoise clipped AWGN run with --model_file model_clipped_noise.pth
  • run with --help to see details on all input parameters

Training

If you want to train your own models you can execute

train_fastdvdnet.py \
	--trainset_dir <path_to_input_mp4s> \
	--valset_dir <path_to_val_sequences> \
	--log_dir logs

NOTES

  • As the dataloader in based on the DALI library, the training sequences must be provided as mp4 files, all under <path_to_input_mp4s>
  • The validation sequences must be stored as image sequences in individual folders under <path_to_val_sequences>
  • run with --help to see details on all input parameters

ABOUT

Copying and distribution of this file, with or without modification, are permitted in any medium without royalty provided the copyright notice and this notice are preserved. This file is offered as-is, without any warranty.

  • Author : Matias Tassano mtassano at gopro dot com
  • Copyright : (C) 2019 Matias Tassano
  • Licence : GPL v3+, see GPLv3.txt

The sequences are Copyright GoPro 2018