{"id":20417433,"url":"https://github.com/hzwer/wacv2024-safa","last_synced_at":"2025-04-05T12:03:53.366Z","repository":{"id":203796461,"uuid":"710228563","full_name":"hzwer/WACV2024-SAFA","owner":"hzwer","description":"WACV2024 - Scale-Adaptive Feature Aggregation for Efficient Space-Time Video Super-Resolution","archived":false,"fork":false,"pushed_at":"2024-11-18T09:33:29.000Z","size":330,"stargazers_count":148,"open_issues_count":1,"forks_count":12,"subscribers_count":8,"default_branch":"main","last_synced_at":"2025-03-29T11:05:40.612Z","etag":null,"topics":["computer-vision","deep-learning","super-resolution","video-processing"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/hzwer.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,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2023-10-26T09:24:29.000Z","updated_at":"2025-03-27T03:58:05.000Z","dependencies_parsed_at":"2023-11-28T09:29:21.902Z","dependency_job_id":"5f2ebf1b-28fc-437b-9672-54e1fae5fb7d","html_url":"https://github.com/hzwer/WACV2024-SAFA","commit_stats":null,"previous_names":["megvii-research/wacv2024-safa","hzwer/wacv2024-safa"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hzwer%2FWACV2024-SAFA","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hzwer%2FWACV2024-SAFA/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hzwer%2FWACV2024-SAFA/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hzwer%2FWACV2024-SAFA/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/hzwer","download_url":"https://codeload.github.com/hzwer/WACV2024-SAFA/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247332560,"owners_count":20921853,"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":["computer-vision","deep-learning","super-resolution","video-processing"],"created_at":"2024-11-15T06:26:24.314Z","updated_at":"2025-04-05T12:03:53.349Z","avatar_url":"https://github.com/hzwer.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Scale-Adaptive Feature Aggregation for Efficient Space-Time Video Super-Resolution\n## [YouTube](https://youtu.be/J4N75OJPGJc) | [Poster](https://drive.google.com/file/d/1kiBWp-qP2lCRIRxbmfOAVtnjleRF0ICq/view?usp=share_link) | [Enhancement Model](https://github.com/hzwer/Practical-RIFE/blob/main/README.md#video-enhancement) | [demo](https://www.youtube.com/watch?v=QII2KQSBBwk) | [中文介绍](https://zhuanlan.zhihu.com/p/668775986)\n## Introduction\nWe want to increase video resolution and frame rates end-to-end (end-to-end STVSR). This project is the implement of [Scale-Adaptive Feature Aggregation for Efficient Space-Time Video Super-Resolution](http://arxiv.org/abs/2310.17294). Our SAFA network outperforms recent state-of-the-art methods such as TMNet and VideoINR by an average improvement of over 0.5dB on PSNR, while requiring less than half the number of parameters and only 1/3 computational costs.\n\nWe have released some dedicated visual effect models for ordinary users. Some insights on multi-scale processing and feature fusion are reflected in RIFE applications, see [Practical-RIFE](https://github.com/hzwer/Practical-RIFE#video-enhancement). \n\nSpace-Time Super-Resolution: \n\n![slomo_origin](https://github.com/megvii-research/WACV2024-SAFA/assets/10103856/aa9710a8-4b23-4c14-adaa-d864431faebd) -\u003e ![slomo](https://github.com/megvii-research/WACV2024-SAFA/assets/10103856/58728e32-ca3b-4cc2-8b8f-b68a7ff9e2ee)\n\n\u003cimg width=\"510\" alt=\"image\" src=\"https://github.com/megvii-research/WACV2024-SAFA/assets/10103856/a243c9e2-243e-4ce6-a5c0-3739d98eb22c\"\u003e\n\n## CLI Usage\n\n### Installation\n\n```\ngit clone git@github.com:megvii-research/WACV2024-SAFA.git\ncd WACV2024-SAFA\npip3 install -r requirements.txt\n```\n\nDownload the pretrained model from [Google Drive](https://drive.google.com/file/d/1PCYRfKwMkymP0V5dmcmGwrKu0lU7xSZ0/view?usp=share_link).\n\n### Run\n\n**Image Interpolation**\n```\npython3 inference_img.py --img demo/i0.png demo/i1.png --exp=3\n```\n(2^3=8X interpolation results)\n\n```\npython3 inference_img.py --img demo/i0.png demo/i1.png --ratio=0.4\n```\n(for an arbitrary timestep)\n\n## Training and Reproduction\nWe use 16 CPUs, 4 GPUs for training: \n```\npython3 -m torch.distributed.launch --nproc_per_node=4 train.py --world_size=4\n```\nThe training scheme is mainly adopted from [RIFE](https://github.com/megvii-research/ECCV2022-RIFE).\n\n![image](https://github.com/hzwer/WACV2024-SAFA/assets/10103856/be0c151d-f8a3-465c-ac78-45d688a31c70) ![image](https://github.com/hzwer/WACV2024-SAFA/assets/10103856/246ba74f-44fd-4001-8375-bb41e9adbcad)\n\n## Recommend\nWe sincerely recommend some related papers:\n\nECCV22 - [Real-Time Intermediate Flow Estimation for Video Frame Interpolation](https://github.com/megvii-research/ECCV2022-RIFE)\n\nCVPR23 - [A Dynamic Multi-Scale Voxel Flow Network for Video Prediction](https://huxiaotaostasy.github.io/DMVFN/)\n\n## Citation\nIf you think this project is helpful, please feel free to leave a star or cite our paper:\n\n```\n@inproceedings{huang2024safa,\n  title={Scale-Adaptive Feature Aggregation for Efficient Space-Time Video Super-Resolution},\n  author={Huang, Zhewei and Huang, Ailin and Hu, Xiaotao and Hu, Chen and Xu, Jun and Zhou, Shuchang},\n  booktitle={Winter Conference on Applications of Computer Vision (WACV)},\n  year={2024}\n}\n```\n## Reference\n\n[RIFE](https://github.com/megvii-research/ECCV2022-RIFE)   [DMVFN](https://huxiaotaostasy.github.io/DMVFN/)   [TMNet](https://github.com/CS-GangXu/TMNet)\n\n[ZoomingSlomo](https://github.com/Mukosame/Zooming-Slow-Mo-CVPR-2020)    [VideoINR](https://github.com/Picsart-AI-Research/VideoINR-Continuous-Space-Time-Super-Resolution)   \n\n![image](https://github.com/megvii-research/WACV2024-SAFA/assets/10103856/d8b92072-bcf7-4d9d-bb27-26c07d85a154)\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhzwer%2Fwacv2024-safa","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhzwer%2Fwacv2024-safa","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhzwer%2Fwacv2024-safa/lists"}