└── README.md /README.md: -------------------------------------------------------------------------------- 1 | # Large-scale Diverse Video (LDV) Dataset 2 | 3 | **Download the latest version: LDV 3.0 (365 videos)** [[Raw videos]](https://drive.google.com/file/d/1MT6oihhUuZDfaMSOZfH5SERxbUkHA4Je/view?usp=drive_link) [[Info]](https://drive.google.com/file/d/1_VTWRhrTKaNvMlR0UH-WZcWJibu1QwLt/view?usp=drive_link) 4 | 5 | The LDV 1.0, LDV 2.0 and LDV 3.0 datasets are established for the Video Enhancement Challenges in [NTIRE Workshop 2021](https://data.vision.ee.ethz.ch/cvl/ntire21/), [NTIRE Workshop 2022](https://data.vision.ee.ethz.ch/cvl/ntire22/) and [AIM Workshop 2022](https://data.vision.ee.ethz.ch/cvl/aim22/). 6 | 7 | If the datasets and the benchmarks are useful for your research, please cite: 8 | ``` 9 | @inproceedings{yang2021dataset, 10 | title={{NTIRE 2021} Challenge on Quality Enhancement of Compressed Video: Dataset and Study}, 11 | author={Ren Yang and Radu Timofte}, 12 | booktitle={IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops}, 13 | year={2021} 14 | } 15 | 16 | @inproceedings{yang2022ntire, 17 | title={{NTIRE 2022} Challenge on Super-Resolution and Quality Enhancement of Compressed Video: Dataset, Methods and Results}, 18 | author={Ren Yang and Radu Timofte and others}, 19 | booktitle={IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops}, 20 | year={2022} 21 | } 22 | 23 | @inproceedings{yang2022aim, 24 | title={{AIM 2022} Challenge on Super-Resolution of Compressed Image and Video: Dataset, Methods and Results}, 25 | author={Ren Yang and Radu Timofte and others}, 26 | booktitle={European Conference on Computer Vision Workshops}, 27 | year={2022} 28 | } 29 | ``` 30 | 31 | Should you have any questions, please feel free to contact: 32 | 33 | Ren Yang @ ETH Zurich, Switzerland 34 | 35 | Email: r.yangchn@gmail.com 36 | 37 | 38 | ## LDV 1.0 (240 videos) 39 | 40 | The proposed LDV 1.0 in NTIRE 2021 contains 240 videos with diverse categories of content, different kinds of motion and various frame-rates. The dataset may be further extended in the future. The details of the proposed LDV 1.0 dataset are discribed in the dataset report: 41 | 42 | > Ren Yang and Radu Timofte, "NTIRE 2021 Challenge on Quality Enhancement of Compressed Video: Dataset and Study", in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2021. [[Paper]](https://arxiv.org/abs/2104.10782) 43 | 44 | ### Download LDV 1.0 45 | 46 | - Training set (200 videos) 47 | [[Raw]](https://drive.google.com/file/d/1E24fD2rGrB5QQWUl30RkczEkaGJlPBH_/view?usp=drive_link) 48 | [[Compressed (fixed QP)]](https://drive.google.com/file/d/1nhfzUHBUzQzhJO8GkpQtNW4vxu4tunNx/view?usp=drive_link) 49 | [[Compressed (fixed bit-rate)]](https://drive.google.com/file/d/1SdIpIqeoQR-MgR0NS06eALL-I_vJjcIj/view?usp=drive_link) 50 | [[Info]](https://drive.google.com/file/d/1SqRKyNxOE7cRb3TZ4JcPpe0r01zrpEq2/view?usp=drive_link) 51 | 52 | - Validation set (20 videos) 53 | [[Raw]](https://drive.google.com/file/d/19bNZhdjvRtbFCmOPzSJowCkjrk1dsecL/view?usp=drive_link) 54 | [[Compressed (fixed QP)]](https://drive.google.com/file/d/1ANB85AI_2ShppyBkOEPvQKBHaS-HlPc1/view?usp=drive_link) 55 | [[Compressed (fixed bit-rate)]](https://drive.google.com/file/d/1XLsytIMWjRYpkBD_CMiE5e9733dC4TvE/view?usp=drive_link) 56 | [[Info]](https://drive.google.com/file/d/1wa4s7AOHpO2iWzK99sJbKaLna4pivmRM/view?usp=drive_link) 57 | 58 | - Test set for Tracks 1 and 2 (10 videos) 59 | [[Raw]](https://drive.google.com/file/d/1W4bDwhQbZpRyWmBlEAJZH6cte5-Blvn8/view?usp=drive_link) 60 | [[Compressed (fixed QP)]](https://drive.google.com/file/d/1mc4LXrseNkpjLe9HfD0ho7g9JPHneKHy/view?usp=drive_link) 61 | [[Compressed (fixed bit-rate)]](https://drive.google.com/file/d/1zeNSwd2D0D6vJ1hmXiOn0C2ug4IJmTca/view?usp=drive_link) 62 | [[Info]](https://drive.google.com/file/d/1gS7Sis0tBvUA-xg96STEPaY2uTGYyGAM/view?usp=drive_link) 63 | 64 | - Test set for Track 3 (10 videos) 65 | [[Raw]](https://drive.google.com/file/d/1bCDGvmE_5ZQq7l4MRtJwyIMKWhj8JVxc/view?usp=drive_link) 66 | [[Compressed (fixed QP)]](https://drive.google.com/file/d/1g08W8nIe6Hv8Mt7LZ0OoNNOPn478s_AR/view?usp=drive_link) 67 | [[Compressed (fixed bit-rate)]](https://drive.google.com/file/d/1Y2pGoDN7yW7NeZ4bZy-UjmXUyaZwaJFd/view?usp=drive_link) 68 | [[Info]](https://drive.google.com/file/d/18Hsm80F9h2_uEVoVoI4GNlZBvEAirYtw/view?usp=drive_link) 69 | 70 | The NTIRE 2021 challenge compresses videos in the **YUV domain** and evaluates results in the **RGB domain**. The following commands can be used to convert the videos to the YUV and RGB domains, respectively. 71 | 72 | ``` 73 | ffmpeg -i xxx.mkv -pix_fmt yuv420p xxx.yuv 74 | ffmpeg -i xxx.mkv ./xxx/%3d.png 75 | ``` 76 | 77 | ## LDV 2.0 (LDV 1.0 + 95 videos = 335 videos) 78 | 79 | The proposed LDV 2.0 in NTIRE 2022 contains 335 videos with diverse categories of content, different kinds of motion and various frame-rates. The details of the proposed LDV 2.0 dataset are discribed in the dataset report: 80 | 81 | > Ren Yang, Radu Timofte, and others, "NTIRE 2022 Challenge on Super-Resolution and Quality Enhancement of Compressed Video: Dataset, Methods and Results", in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022. [[Paper]](https://openaccess.thecvf.com/content/CVPR2022W/NTIRE/papers/Yang_NTIRE_2022_Challenge_on_Super-Resolution_and_Quality_Enhancement_of_Compressed_CVPRW_2022_paper.pdf) 82 | > 83 | 84 | ### Download LDV 2.0 85 | 86 | - **The whole dataset (335 videos)** [[Raw videos]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/LDV2.zip) [[Info]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/data_LDV2.xlsx) 87 | 88 | - **Track 1** 89 | - Training set (LDV 1.0, 240 videos): 90 | [[Raw]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/train/train_gt.zip) 91 | [[Compressed (Fixed QP)]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/train/train_QP37.zip) 92 | [[Info]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/train/data_train_1.xlsx) 93 | - Validation set (15 videos): 94 | [[Raw]](https://drive.google.com/file/d/1rX7_asAGa4xOVnGc2l9QYoYyVa7yBFSM/view?usp=drive_link) 95 | [[Compressed (Fixed QP)]](https://drive.google.com/file/d/1hvGPYWI6U6XXLKQTBGk28KDyZ6CM4FH3/view?usp=drive_link) 96 | [[Info]](https://drive.google.com/file/d/1FKQk4P6d56676XHRaaSaFq01fdpBgkec/view?usp=drive_link) 97 | - Test set (15 videos): 98 | [[Raw]](https://drive.google.com/file/d/1yF7b5NG-tgLPkbPTeogt1nI3tXtN6hAe/view?usp=drive_link) 99 | [[Compressed (Fixed QP)]](https://drive.google.com/file/d/1LcXCaDRccKGFg8fUdB23iX9LLR5s3D4q/view?usp=drive_link) 100 | [[Info]](https://drive.google.com/file/d/1_cLeRy64rgB7SK7aMEcmLTYkg7NoOoTA/view?usp=drive_link) 101 | 102 | - **Track 2** (The sizes of raw videos are cropped to the multiples of 16) 103 | - Training set (LDV 1.0, 240 videos): 104 | [[Raw]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/train/train_down2_gt.zip) 105 | [[Compressed (Fixed QP)]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/train/train_down2_QP37.zip) 106 | [[Info]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/train/data_train_2.xlsx) 107 | - Validation set (15 videos): 108 | [[Raw]](https://drive.google.com/file/d/1GXfakFwFgb2G8f1B7g9y1D2tV5MAXzpW/view?usp=drive_link) 109 | [[Compressed (Fixed QP)]](https://drive.google.com/file/d/1F5sMpu-Y8KHhz23UKZurJ-ewiwhdQBqR/view?usp=drive_link) 110 | [[Info]](https://drive.google.com/file/d/1tGw_ab_3kDi3iDFCjmP0czbYV-qtG2Ml/view?usp=drive_link) 111 | - Test set (15 videos): 112 | [[Raw]](https://drive.google.com/file/d/1RMM6fXNfDmPqsz6dZ95yPVZ2WgYKbGzA/view?usp=drive_link) 113 | [[Compressed (Fixed QP)]](https://drive.google.com/file/d/16QYY0rZ8a3avSJgI3Kv55ZMNrl__WtmZ/view?usp=drive_link) 114 | [[Info]](https://drive.google.com/file/d/1-ps3uxXuJWsavvTnDhVu7QFOQ_7A2HbM/view?usp=drive_link) 115 | 116 | - **Track 3** (The sizes of raw videos are cropped to the multiples of 64) 117 | - Training set (LDV 1.0, 240 videos): 118 | [[Raw]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/train/train_down4_gt.zip) 119 | [[Compressed (Fixed QP)]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/train/train_down4_QP37.zip) 120 | [[Info]](https://data.vision.ee.ethz.ch/reyang/NTIRE2022/train/data_train_3.xlsx) 121 | - Validation set (15 videos): 122 | [[Raw]](https://drive.google.com/file/d/1pIn_dcH4yvFSUBT08D4QK18guWaM6wnd/view?usp=drive_link) 123 | [[Compressed (Fixed QP)]](https://drive.google.com/file/d/1vT8jVmyUeJGLlWiEBBCU-F8PKBZ_I-55/view?usp=drive_link) 124 | [[Info]](https://drive.google.com/file/d/1UcW_E72tDRpsThDK5HA1PChtnVGUz6F-/view?usp=drive_link) 125 | - Test set (15 videos): 126 | [[Raw]](https://drive.google.com/file/d/1dqAoS5nCuBMlNoMcHqaMPXoSSOSZehC4/view?usp=drive_link) 127 | [[Compressed (Fixed QP)]](https://drive.google.com/file/d/1P5juq-cXGy5-YRQU8BVmJWgzB5M29JRw/view?usp=drive_link) 128 | [[Info]](https://drive.google.com/file/d/1jkAwh4L93huWUGJ5Gy2NVJEJPJ_unIV8/view?usp=drive_link) 129 | 130 | The NTIRE 2022 challenge compresses videos in the **YUV domain** and evaluates results in the **RGB domain**. The following commands can be used to convert the videos to the YUV and RGB domains, respectively. 131 | 132 | ``` 133 | ffmpeg -i xxx.mkv -pix_fmt yuv420p xxx.yuv 134 | ffmpeg -i xxx.mkv ./xxx/%3d.png 135 | ``` 136 | 137 | ## LDV 3.0 (LDV 2.0 + 30 videos = 365 videos) 138 | 139 | The proposed LDV 3.0 in AIM 2022 contains 365 videos with diverse categories of content, different kinds of motion and various frame-rates. The details of the proposed LDV 3.0 dataset are discribed in the dataset report: 140 | 141 | > Ren Yang, Radu Timofte, and others, "AIM 2022 Challenge on Super-Resolution of Compressed Image and Video: Dataset, Methods and Results", in European Conference on Computer Vision Workshops, 2022. [[Paper]](https://arxiv.org/pdf/2208.11184.pdf) 142 | > 143 | 144 | ### Download LDV 3.0 145 | 146 | - **The whole dataset (365 videos)** [[Raw videos]](https://drive.google.com/file/d/1MT6oihhUuZDfaMSOZfH5SERxbUkHA4Je/view?usp=drive_link) [[Info]](https://drive.google.com/file/d/1_VTWRhrTKaNvMlR0UH-WZcWJibu1QwLt/view?usp=drive_link) 147 | 148 | - **Track 2** (The sizes of raw videos are cropped to the multiples of 64) 149 | - Validation set (15 videos): 150 | [[Raw]](https://drive.google.com/file/d/1PE3yHukHm850LaqAY3QZNpmMdjiQ31Z5/view?usp=drive_link) 151 | [[Compressed (Fixed QP)]](https://drive.google.com/file/d/1E3rTScRq58xNXp4JR7Xw_Gs7G9MTY8Af/view?usp=drive_link) 152 | [[Info]](https://drive.google.com/file/d/1mufD_-ceg8wLkaVnoTWy2d09cIGnVwdI/view?usp=drive_link) 153 | - Test set (15 videos): 154 | [[Raw]](https://drive.google.com/file/d/1tCLh8KAK7JcwfwNWsytk6KYpG-RgWGqq/view?usp=drive_link) 155 | [[Compressed (Fixed QP)]](https://drive.google.com/file/d/1RgZRymK3i4TMxHSSZbk48p9RkF8uULjl/view?usp=drive_link) 156 | [[Info]](https://drive.google.com/file/d/1IAsho89oY8j3_uV0yx2Pi2kG5MC5X3bv/view?usp=drive_link) 157 | 158 | The AIM 2022 challenge compresses videos in the **YUV domain** and evaluates results in the **RGB domain**. The following commands can be used to convert the videos to the YUV and RGB domains, respectively. 159 | 160 | ``` 161 | ffmpeg -i xxx.mkv -pix_fmt yuv420p xxx.yuv 162 | ffmpeg -i xxx.mkv ./xxx/%3d.png 163 | ``` 164 | --------------------------------------------------------------------------------