{"id":13543557,"url":"https://github.com/m-tassano/dvdnet","last_synced_at":"2025-04-02T13:30:40.531Z","repository":{"id":51216661,"uuid":"168368325","full_name":"m-tassano/dvdnet","owner":"m-tassano","description":"DVDnet: A Simple and Fast Network for Deep Video Denoising","archived":false,"fork":false,"pushed_at":"2021-09-11T10:12:59.000Z","size":4922,"stargazers_count":79,"open_issues_count":0,"forks_count":18,"subscribers_count":6,"default_branch":"master","last_synced_at":"2024-11-03T10:33:03.228Z","etag":null,"topics":["cnn","cnn-architecture","deep-learning","denoising","python3","pytorch","video"],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"gpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/m-tassano.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2019-01-30T15:43:42.000Z","updated_at":"2024-09-27T13:08:13.000Z","dependencies_parsed_at":"2022-08-12T23:00:12.543Z","dependency_job_id":null,"html_url":"https://github.com/m-tassano/dvdnet","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/m-tassano%2Fdvdnet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/m-tassano%2Fdvdnet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/m-tassano%2Fdvdnet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/m-tassano%2Fdvdnet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/m-tassano","download_url":"https://codeload.github.com/m-tassano/dvdnet/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":246823539,"owners_count":20839744,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cnn","cnn-architecture","deep-learning","denoising","python3","pytorch","video"],"created_at":"2024-08-01T11:00:32.928Z","updated_at":"2025-04-02T13:30:37.136Z","avatar_url":"https://github.com/m-tassano.png","language":"Python","funding_links":[],"categories":["Denoising Algorithms"],"sub_categories":[],"readme":"[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-davis-sigma30)](https://paperswithcode.com/sota/video-denoising-on-davis-sigma30?p=dvdnet-a-fast-network-for-deep-video)\n[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-davis-sigma40)](https://paperswithcode.com/sota/video-denoising-on-davis-sigma40?p=dvdnet-a-fast-network-for-deep-video)\n[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-davis-sigma50)](https://paperswithcode.com/sota/video-denoising-on-davis-sigma50?p=dvdnet-a-fast-network-for-deep-video)\n[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-set8-sigma20)](https://paperswithcode.com/sota/video-denoising-on-set8-sigma20?p=dvdnet-a-fast-network-for-deep-video)\n[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-set8-sigma30)](https://paperswithcode.com/sota/video-denoising-on-set8-sigma30?p=dvdnet-a-fast-network-for-deep-video)\n[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-set8-sigma40)](https://paperswithcode.com/sota/video-denoising-on-set8-sigma40?p=dvdnet-a-fast-network-for-deep-video)\n[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-set8-sigma50)](https://paperswithcode.com/sota/video-denoising-on-set8-sigma50?p=dvdnet-a-fast-network-for-deep-video)\n[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-davis-sigma10)](https://paperswithcode.com/sota/video-denoising-on-davis-sigma10?p=dvdnet-a-fast-network-for-deep-video)\n[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-davis-sigma20)](https://paperswithcode.com/sota/video-denoising-on-davis-sigma20?p=dvdnet-a-fast-network-for-deep-video)\n[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/dvdnet-a-fast-network-for-deep-video/video-denoising-on-set8-sigma10)](https://paperswithcode.com/sota/video-denoising-on-set8-sigma10?p=dvdnet-a-fast-network-for-deep-video)\n\n# [DVDnet](https://hal.archives-ouvertes.fr/hal-02147604)\nA state-of-the-art, simple and fast network for Deep Video Denoising\n\n**NEW: a state-of-the-art algorithm for video denoising without motion compensation [FastDVDnet](https://github.com/m-tassano/fastdvdnet)**\n\n## Overview\n\nThis source code provides a PyTorch implementation of DVDnet image denoising, as in \nTassano, Matias and Delon, Julie and Veit, Thomas. \"DVDnet: A Fast Network for Deep Video Denoising\", IEEE ICIP 2019, arXiv preprint arXiv:1906.11890 (2019).\n\n## Video examples\n You can download several denoised sequences with our algorithm and other methods [here](https://www.dropbox.com/sh/gccey7wuxiqr104/AAC_v6kb3fMYxMHBc6wcqu17a?dl=0 \"DVDnet denoised sequences\")\n \n## User Guide\n\nThe code as is runs in Python +3.6 with the following dependencies:\n### Dependencies\n* [PyTorch v1.0.0](http://pytorch.org/)\n* [scikit-image](http://scikit-image.org/)\n* [numpy](https://www.numpy.org/)\n* [OpenCV](https://pypi.org/project/opencv-python/)\n\n## Usage\n\nIf you want to denoise an image sequence using the pretrained models\nfound under the *models* folder you can execute\n\n```\npython test_dvdnet.py \\\n        --test_path \u003cpath_to_input_sequence\u003e \\\n\t--save_path results \\\n        --noise_sigma 25 \\\n```\n\n**NOTES**\n* The image sequence should be stored under \u003cpath_to_input_sequence\u003e\n* Models have been trained for values of noise in [5, 55]\n* run with *--no_gpu* to run on CPU instead of GPU\n* run with *--save_noisy* to save noisy frames\n* set *max_num_fr_per_seq* to set the max number of frames to load per sequence\n* run with *--help* to see details on all input parameters\n\n## Comparison of PSNRs\nTwo different testsets were used for benchmarking our method: the DAVIS-test testset, and Set8, which is composed of 4 color sequences from the [Derf’s Test Media collection](https://media.xiph.org/video/derf) and 4 color sequences captured with a GoPro camera. The DAVIS set contains 30 color sequences of resolution 854 x 480. The sequences of Set8 have been downscaled to a resolution of 960 x 540. In all cases, sequences were limited to a maximum of 85 frames. We used the DeepFlow algorithm to compute flow maps for DVDnet and VNLB. For Neat Video, the automatic noise profiling settings were used.\n\nNote: values shown are the average for all sequences in the testset, the PNSR of a sequence is computed as the average of the PSNRs of each frame.\n\n### PSNRs Set8 testset\n| Noise std dev | DVDNet | VNLB [1] | V-BM4D [2] | Neat Video [3] |\n|---|---|---|---|---|\n| 10 | 36.08 | **37.26** | 36.05 | 35.67 | \n| 20 | 33.49 | **33.72** | 32.19 | 31.69 | \n| 30 | **31.79** | 31.74 | 30.00 | 28.84 | \n| 40 | **30.55** | 30.39 | 28.48 | 26.36 | \n| 50 | **29.56** | 29.24 | 27.33 | 25.46 | \n\n### PSNRs DAVIS testset\n| Noise std dev | DVDNet | VNLB | V-BM4D |\n|--|--|--|--|\n| 10 | 38.13 | **38.85** | 37.58 |\n| 20 | **35.70** | 35.68 | 33.88 |\n| 30 | **34.08** | 33.73 | 31.65 |\n| 40 | **32.86** | 32.32 | 30.05 |\n| 50 | **31.85** | 31.13 | 28.80 |\n\n## ABOUT\n\nCopying and distribution of this file, with or without modification,\nare permitted in any medium without royalty provided the copyright\nnotice and this notice are preserved. This file is offered as-is,\nwithout any warranty.\n\n* Author    : Matias Tassano `mtassano at gopro dot com`\n* Copyright : (C) 2019 Matias Tassano\n* Licence   : GPL v3+, see GPLv3.txt\n\nThe sequences are Copyright GoPro 2018\n\n## References\n\n[1] P. Arias and J.-M. Morel, “Video denoising via empirical Bayesian estimation of space-time patches,” Journal of Mathematical Imaging and Vision, vol. 60, no. 1, pp. 70—-93, 2018\n\n[2] M. Maggioni, G. Boracchi, A. Foi, K. Egiazarian, “Video denoising, deblocking, and enhancement through separable 4-D nonlocal spatiotemporal transforms,” IEEE Trans. IP, vol. 21, no. 9, pp. 3952–3966, 2012.\n\n[3] ABSoft, “Neat Video,” https://www.neatvideo.com, 1999–2019.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fm-tassano%2Fdvdnet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fm-tassano%2Fdvdnet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fm-tassano%2Fdvdnet/lists"}