Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
17,713 datasets
Unpaired images of horses and zebras, originally part of the CycleGAN benchmark suite. It is derived from the research paper 'Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks' (arXiv:1703.10593). The dataset is hosted by the 'huggan' organization and was last updated in April 2022.
ImageNet Sketch Data is a collection of sketch images created by nateraw and hosted on Hugging Face, last updated in May 2022. It contains sketch-style images corresponding to the ImageNet classification challenge categories. The dataset is designed for evaluating model performance on out-of-distribution, non-photographic data.
Part of the CycleGAN datasets collection for unpaired image-to-image translation. It contains images of Monet paintings and photographs, as referenced in the 2017 CycleGAN research paper by Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A. Efros. The specific row count, column structure, and file formats are not provided in the input.
Tiny ImageNet is a smaller-scale version of the ImageNet dataset, containing 100,000 images across 200 object classes. It was created by yzhou992 and uploaded to Hugging Face in May 2022. This dataset serves as a more manageable benchmark for training and evaluating image classification models.
A collection of unpaired images of apples and oranges, originally part of the CycleGAN benchmark suite. It was created by Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A. Efros for research on unpaired image-to-image translation. The dataset is hosted by the 'huggan' organization and was last updated in April 2022.
Comprising unpaired images of Van Gogh paintings and photographs, originally part of the CycleGAN benchmark. It is used for training unpaired image-to-image translation models. The dataset is associated with the 2017 research paper on Cycle-Consistent Adversarial Networks.
Encompassing unpaired images of paintings in the style of Paul Cezanne and photographs. It is part of the CycleGAN datasets collection for image-to-image translation research, originally published by researchers at UC Berkeley.
Part of the CycleGAN datasets for unpaired image-to-image translation. It contains cat images used in the original CycleGAN research paper published in 2017. The specific row count, column structure, and image count are not detailed in the provided input.
Encompassing examples of user intentions and greetings in Sheng, a Swahili-based slang. The data includes phrases for common interactions like 'Greetings' and 'Affirm'. The specific number of rows, columns, and sample data are unavailable.
21,551 cropped anime face images sourced from getchu.com and processed via the lbpcascade_animeface detection algorithm. All images are standardized to a 64x64 pixel resolution to facilitate rapid model training and experimentation.
Functioning as a version of the CORD dataset prepared for the LayoutLMv3 model, focusing on document understanding tasks. It was authored by 'nielsr' and last updated in May 2022. Specific details on row count, column structure, and file formats are unavailable.
15,103 images across Cityscapes and PASCAL VOC benchmarks annotated with part-level segmentation masks for semantic object classes. The dataset provides a unified framework for reading, processing, and evaluating hierarchical scene labels as introduced in the CVPR 2021 paper.
200 high-resolution video sequences containing over 300,000 frames of synchronized RGB and depth data for object tracking. The dataset covers diverse indoor and outdoor environments with manual bounding box annotations provided for every frame.
Imagenette2 320 is a curated subset of the ImageNet dataset, created by user johnowhitaker. It contains 10 easily classified classes from the original ImageNet hierarchy, with images resized to 320 pixels. The dataset was last updated on Hugging Face in May 2022.
10 examples from the segments/sidewalk-semantic dataset, consisting of 10 images with corresponding ground-truth segmentation maps. It was published by Hugging Face in April 2022.
Comprising a filtered subset of the iNaturalist butterfly images. Images were selected using CLIP to compare each image with a text description of a good image. The original dataset is from the huggan/inat_butterflies repository.
Video Understanding is a dataset hosted on HuggingFace by user filwsyl, last updated in May 2022. The dataset is intended for tasks related to interpreting video content. Specific details on the number of videos, their format, or content are not provided in the available metadata.
Serving as associated with the Pix2Pix model for image-to-image translation. It contains paired images for translating scenes from nighttime to daytime conditions. The dataset was created by the HugGAN organization and was last updated in April 2022.
Used for image-to-image translation tasks, as described in the 2017 CVPR paper by Isola et al. It contains pairs of images for training conditional adversarial networks. The specific row count, column structure, and image count are not provided in the input.
A collection of 13,427 camera images with approximately 24,000 annotated traffic lights. The annotations include bounding boxes and the active light state for each traffic light. Images are provided as raw 12-bit HDR and reconstructed 8-bit RGB formats.