Datasets:

ROSCOSMOS
/

Sports_Cars

unique_id string	width int32	height int32	original_file_format string	image_mode_on_disk string
img_00001_17c6	4,608	3,072	JPEG	RGB
img_00002_319c	5,304	7,952	JPEG	RGB
img_00003_cc6c	4,000	6,000	JPEG	RGB
img_00004_da0a	3,743	4,679	JPEG	RGB
img_00005_a4e1	4,278	5,348	JPEG	RGB
img_00006_bfe6	5,744	3,829	JPEG	RGB
img_00007_096e	6,000	4,000	JPEG	RGB
img_00008_3c43	3,526	4,407	JPEG	RGB
img_00009_0521	6,720	4,480	JPEG	RGB
img_00010_9086	3,717	6,608	JPEG	RGB
img_00011_a3d1	6,000	4,000	JPEG	RGB
img_00012_9fc5	5,989	3,993	JPEG	RGB
img_00013_ea73	6,702	4,127	JPEG	RGB
img_00014_5963	5,184	3,456	JPEG	RGB
img_00015_b70e	7,008	4,672	JPEG	RGB
img_00016_9109	6,000	4,000	JPEG	RGB
img_00017_4920	3,600	5,400	JPEG	RGB
img_00018_6c23	5,737	3,825	JPEG	RGB
img_00019_89e7	6,000	4,000	JPEG	RGB
img_00020_3fa1	6,000	4,000	JPEG	RGB
img_00021_5f92	4,102	6,154	JPEG	RGB
img_00022_381b	6,000	4,000	JPEG	RGB
img_00023_cf0c	6,000	4,000	JPEG	RGB
img_00024_73fa	4,160	6,240	JPEG	RGB
img_00025_e120	6,362	3,860	JPEG	RGB
img_00026_8114	3,845	5,768	JPEG	RGB
img_00027_a176	4,000	6,000	JPEG	RGB
img_00028_7709	6,000	4,000	JPEG	RGB
img_00029_ac0d	4,071	5,089	JPEG	RGB
img_00030_30bb	2,884	4,326	JPEG	RGB
img_00031_9e56	4,074	5,093	JPEG	RGB
img_00032_58ae	4,000	6,000	JPEG	RGB
img_00033_d702	5,184	3,456	JPEG	RGB
img_00034_c77a	5,774	3,849	JPEG	RGB
img_00035_3ff7	6,005	4,003	JPEG	RGB
img_00036_e5fc	3,830	2,713	JPEG	RGB
img_00037_459b	3,072	4,608	JPEG	RGB
img_00038_2d53	4,000	6,000	JPEG	RGB
img_00039_3e89	3,896	5,883	JPEG	RGB
img_00040_a2be	6,000	4,000	JPEG	RGB
img_00041_6ae1	4,618	3,079	JPEG	RGB
img_00042_dff5	6,720	4,480	JPEG	RGB
img_00043_97c9	6,000	4,000	JPEG	RGB
img_00044_f5f4	4,136	2,758	JPEG	RGB
img_00045_aa63	4,160	6,240	JPEG	RGB
img_00046_d580	3,456	5,184	JPEG	RGB
img_00047_9d55	5,948	3,965	JPEG	RGB
img_00048_7625	6,000	4,000	JPEG	RGB
img_00049_85d2	3,875	4,844	JPEG	RGB
img_00050_186b	4,145	6,217	JPEG	RGB
img_00051_c481	6,000	4,000	JPEG	RGB
img_00052_a616	4,393	6,034	JPEG	RGB
img_00053_94a7	6,000	4,000	JPEG	RGB
img_00054_abac	4,480	6,720	JPEG	RGB
img_00055_9a8c	6,000	4,000	JPEG	RGB
img_00056_f7c2	3,468	5,202	JPEG	RGB
img_00057_a558	3,743	4,679	JPEG	RGB
img_00058_423a	6,000	4,000	JPEG	RGB
img_00059_afcf	6,960	4,640	JPEG	RGB
img_00060_052d	4,000	6,000	JPEG	RGB
img_00061_ba8e	3,867	2,578	JPEG	RGB
img_00062_c7f5	5,796	3,864	JPEG	RGB
img_00063_b3a5	6,860	4,573	JPEG	RGB
img_00064_1a94	6,720	4,480	JPEG	RGB
img_00065_5b9c	5,067	7,703	JPEG	RGB
img_00066_1532	4,442	6,603	JPEG	RGB
img_00067_7f53	7,360	4,912	JPEG	RGB
img_00068_ff0e	7,360	4,912	JPEG	RGB
img_00069_b31b	7,360	4,912	JPEG	RGB
img_00070_05c2	7,680	4,320	JPEG	RGB
img_00071_9083	5,074	7,352	JPEG	RGB
img_00072_60c0	8,192	5,461	JPEG	RGB
img_00073_69a3	7,680	4,320	JPEG	RGB
img_00074_6b35	4,648	7,617	JPEG	RGB
img_00075_0e2d	5,304	7,952	JPEG	RGB
img_00076_e658	8,000	5,331	JPEG	RGB
img_00077_0869	7,000	3,500	JPEG	RGB
img_00078_7136	5,184	3,456	JPEG	RGB
img_00079_06b0	4,500	2,320	JPEG	RGB
img_00080_3bed	8,368	5,584	JPEG	RGB
img_00081_6c5a	5,472	3,648	JPEG	RGB
img_00082_c607	5,184	3,456	JPEG	RGB
img_00083_2163	10,000	4,651	JPEG	RGB
img_00084_f68a	4,600	2,400	JPEG	RGB
img_00085_71dc	5,200	2,734	JPEG	RGB
img_00086_f500	7,166	3,822	JPEG	RGB
img_00087_d91a	6,889	3,954	JPEG	RGB
img_00088_59eb	6,656	3,744	JPEG	RGB
img_00089_818b	4,800	2,092	JPEG	RGB
img_00090_1d6b	5,491	2,389	JPEG	RGB
img_00091_5a27	3,816	5,724	JPEG	RGB
img_00092_8ec0	5,367	2,427	JPEG	RGB
img_00093_644e	2,848	4,272	JPEG	RGB
img_00094_d9e6	4,218	3,000	JPEG	RGB
img_00095_e70d	5,039	2,440	JPEG	RGB
img_00096_7797	5,472	3,648	JPEG	RGB
img_00097_e80b	6,000	4,000	JPEG	RGB
img_00098_5d4c	6,000	4,000	JPEG	RGB
img_00099_e69c	3,648	5,472	JPEG	RGB
img_00100_390c	3,648	5,472	JPEG	RGB

End of preview. Expand in Data Studio

Sports Cars

High resolution image subset from the Aesthetic-Train-V2 dataset, contains a mix of modified street cars, high performance / super cars from various manufacturers.

Dataset Details

Curator: Roscosmos
Version: 1.0.0
Total Images: 600
Average Image Size (on disk): ~5.1 MB compressed
Primary Content: Sports Cars
Standardization: All images are standardized to RGB mode and saved at 95% quality for consistency.

Dataset Creation & Provenance

1. Original Master Dataset

This dataset is a subset derived from: zhang0jhon/Aesthetic-Train-V2

Link: https://huggingface.co/datasets/zhang0jhon/Aesthetic-Train-V2
Providence: Large-scale, high-resolution image dataset, refer to its original dataset card for full details.
Original License: MIT

2. Iterative Curation Methodology

CLIP retrieval / manual curation.

Dataset Structure & Content

train split: Contains the full, high-resolution image data and associated metadata. This is the recommended split for model training and full data analysis.

Each example (row) in both splits contains the following fields:

image: The actual image data. In the train split, this is full-resolution.
unique_id: A unique identifier assigned to each image.
width: The width of the image in pixels (from the full-resolution image).
height: The height of the image in pixels (from the full-resolution image).

Citation

@inproceedings{zhang2025diffusion4k,
    title={Diffusion-4K: Ultra-High-Resolution Image Synthesis with Latent Diffusion Models},
    author={Zhang, Jinjin and Huang, Qiuyu and Liu, Junjie and Guo, Xiefan and Huang, Di},
    year={2025},
    booktitle={IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
}
@misc{zhang2025ultrahighresolutionimagesynthesis,
    title={Ultra-High-Resolution Image Synthesis: Data, Method and Evaluation},
    author={Zhang, Jinjin and Huang, Qiuyu and Liu, Junjie and Guo, Xiefan and Huang, Di},
    year={2025},
    note={arXiv:2506.01331},
}

Disclaimer and Bias Considerations

Please consider any inherent biases from the original dataset and those potentially introduced by the automated filtering (e.g., CLIP's biases) and manual curation process.

Contact

N/A

Downloads last month: 29

Size of downloaded dataset files:

3.3 GB

Size of the auto-converted Parquet files:

3.29 GB

Number of rows:

600