datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
fish-vista
Dataset Card for Fish-Visual Trait Analysis (Fish-Vista)
Note that the '</Use this dataset>' option will only load the CSV files. To download the entire dataset, including all processed images and segmentation annotations, refer to Instructions for downloading dataset and images.
See Example Code to Use the Segmentation Dataset
Figure 1. A schematic representation of the different tasks in Fish-Vista Dataset.
Instructions for downloading dataset… See the full description on the dataset page: https://huggingface.co/datasets/imageomics/fish-vista.image-as-an-imu-finetuning
Image as an IMU: Real-world Finetuning Dataset
Official real-world finetuning dataset from Image as an IMU: Estimating Camera Motion from a Single Motion-Blurred Image (ICCV 2025 Oral).
[arXiv] [Webpage] [GitHub]
PIXL, University of Oxford
Jerred Chen, Ronald Clark
Dataset Details
This dataset consists of 32 sequences of real-world motion-blurred videos in various indoor scenes, captured using the iPhone 13 camera.
dataset_train_real-world.csv and… See the full description on the dataset page: https://huggingface.co/datasets/jerredchen00/image-as-an-imu-finetuning.questFish2024
Dataset Card for QUEST Fish 2024
Images collected by teachers during a QUEST workshop. In 2024, the images were of fish collected from bodies of water near Princeton University.
Dataset Details
Dataset Structure
/dataset/
<folder>/
<img_id 1>.png
<img_id 2>.png
...
<img_id n>.png
...
<img_id 1>.png
<img_id 2>.png
...
<img_id n>.png
fieldData2024.csv
Data Instances… See the full description on the dataset page: https://huggingface.co/datasets/imageomics/questFish2024.Heliconius-Collection_Cambridge-Butterfly
Dataset Card for Heliconius Collection (Cambridge Butterfly)
Dataset Description
Dataset Summary
Subset of the collection records from Chris Jiggins' research group at the University of Cambridge, collection covers nearly 20 years of field studies.
This subset contains approximately 36,189 RGB images of 11,962 specimens (29,134 images of 10,086 specimens across all Heliconius). Many records have both images and locality data.
Most images were… See the full description on the dataset page: https://huggingface.co/datasets/imageomics/Heliconius-Collection_Cambridge-Butterfly.Whole_Slide_Imageskabr-methodology
Dataset Card for kabr-tools Methodology Dataset
Dataset Details
A curated collection of CSV and XML files describing time-budget data, focal observations, scan samples, and object-detection annotations for African ungulates—including Grevy’s zebras, plains zebras, and giraffes—recorded both from the ground and from drones. This dataset complements the original KABR Mini-Scene Dataset, providing ground-based sampling to correspond with a subset of the published… See the full description on the dataset page: https://huggingface.co/datasets/imageomics/kabr-methodology.char-sim-data
Dataset Card for Character Similarity Dataset
Dataset Details
The Character Similarity Dataset is a collection of textual trait descriptions along with the corresponding ontology based similarity measures between trait description pairs. The distance is estimated using the Phenoscape Knowledgebase as the ontology. The Knowledgebase is built upon a number of OBO ontologies, most importantly the Uberon anatomy ontology.
The Character Similarity Dataset is a collection of… See the full description on the dataset page: https://huggingface.co/datasets/imageomics/char-sim-data.wikipedia-image-requests
Daily image-request counts for Wikipedia article images
Daily counts of how often each of ~1.19 million Wikipedia article images was requested from
Wikimedia's image servers, attributed to the wiki whose page the request came from, together with
the article each image appears on.
406,263,728 daily observations covering 1,186,150 image files across 534,067 articles in
nine Wikipedias, from 2025-09-01 to 2026-09-05 (370 days).
Every file in the set appears on exactly one… See the full description on the dataset page: https://huggingface.co/datasets/lgelauff/wikipedia-image-requests.imagebench
ImageBench — 50 Text-to-Image Models Judged by VLMs on 192 Prompts
Reproducibility dataset for imagebench.ai: the 192-prompt V1.2 benchmark, per-(model, prompt) VLM verdicts, and per-model aggregate scores for 50 text-to-image models.
Live leaderboard + every generated image: https://imagebench.ai
Methodology: https://imagebench.ai/methodology-v1
Reproducibility repo: https://github.com/dh7/image-bench-ai
What's in this dataset
File
Rows
Description… See the full description on the dataset page: https://huggingface.co/datasets/dh7/imagebench.ImageHeterogeneity
Replication Data for: Image-based Treatment Effect Heterogeneity
Details: UgandaDataProcessed.csv contains individual-level data from the YOP experiment. In the dataset, geo_long and geo_lat refer to the approximate geo-referenced long/lat of experimental units. The variable, geo_long_lat_key, refers to the image key associated with each location. Experimental outcomes are stored in Yobs. Treatment variable is stored in Wobs. See the tutorial for more information.
UgandaGeoKeyMat.csv… See the full description on the dataset page: https://huggingface.co/datasets/cjerzak/ImageHeterogeneity.imagenet-12k-metadata
ImageNet-12k Split Metadata
Metadata files defining the splits for ImageNet-12k subset of fall11_whole.tar (2011 ImageNet full release) used in some timm models (see dataset building code in https://github.com/rwightman/imagenet-12k).
ambient-o-clip-iqa-patches-imagenet
Ambient Diffusion Omni (Ambient-o): Training Good Models with Bad Data
Dataset Description
Ambient Diffusion Omni (Ambient-o) is a framework for using low-quality, synthetic, and out-of-distribution images to improve the quality of diffusion models. Unlike traditional approaches that rely on highly curated datasets, Ambient-o extracts valuable signal from all available images during training, including data typically discarded as "low-quality."
This dataset card is for… See the full description on the dataset page: https://huggingface.co/datasets/adrianrm/ambient-o-clip-iqa-patches-imagenet.Latent-Resonance-AI-Image-Forensics-Benchmark-N1000
Latent Resonance: SOTA Large-Scale AI Image Forensics Benchmark (N=1,000)
Author: Debdip Bandyopadhyay (Independent AI Researcher, Kolkata, India; M.Tech, IIT Jodhpur, AI & Data Science)Preprint & Paper: Latent Resonance: Zero-Shot Autoencoder Inversion and Azimuthal Spectral Forensics for Diffusion Image Attribution (IEEE Flagship / CERN Zenodo 2026)
1. Executive Summary & Diagnostic Suite
This repository contains the complete empirical evaluation records… See the full description on the dataset page: https://huggingface.co/datasets/DebdipCS/Latent-Resonance-AI-Image-Forensics-Benchmark-N1000.gpt-image-edit-benchmark-results
GPT-Image-Edit — Benchmark Results
This repository contains evaluation results of GPT-Image-Edit across four standard image-editing benchmarks. All scores were computed using the official evaluation scripts provided by each benchmark.
📊 Benchmarks
Benchmark
Metrics
Folder
GEdit-EN
12 editing categories + Avg
gedit/
Complex-Edit
IF, IP, PQ, Overall
complex_edit/
ImgEdit-Full
10 editing operations + Overall
imgedit/
OmniContext
Contextual edit scores… See the full description on the dataset page: https://huggingface.co/datasets/UCSC-VLAA/gpt-image-edit-benchmark-results.dosis-radiacion-estudios-imagen-es
Dosis de radiación en estudios de imagen — referencia en español
Tabla de referencia con la dosis efectiva aproximada de 27 estudios de imagen diagnóstica,
en español, con equivalencias comprensibles para pacientes.
Existe buena documentación sobre dosis de radiación en inglés, pero muy poca en español
estructurada y citable. Este conjunto de datos traduce y normaliza la tabla de referencia
pública de RSNA/ACR (RadiologyInfo.org), conservando la cifra original de cada estudio y… See the full description on the dataset page: https://huggingface.co/datasets/NODARISHUB/dosis-radiacion-estudios-imagen-es.ImageNet-CJ
JPEG Re-encoding Confound Control Dataset
A controlled-experiment dataset that isolates one acknowledged-but-unmeasured confound in
ImageNet-C. Hendrycks & Dietterich (Benchmarking Neural Network Robustness to Common
Corruptions and Perturbations, ICLR 2019, arXiv:1903.12261)
save every corrupted image as a lightly compressed JPEG. The benchmark therefore never measures a
corruption c applied to an image x in isolation — it measures JPEG(c(x)). This dataset lets
you quantify how… See the full description on the dataset page: https://huggingface.co/datasets/atharvadagaonkar/ImageNet-CJ.chatinterface_with_image_csv
Dataset Card for Dataset Name
Dataset Summary
[More Information Needed]
Supported Tasks and Leaderboards
[More Information Needed]
Languages
[More Information Needed]
Dataset Structure
Data Instances
[More Information Needed]
Data Fields
[More Information Needed]
Data Splits
[More Information Needed]
Dataset Creation
Curation Rationale
[More Information Needed]
Source Data… See the full description on the dataset page: https://huggingface.co/datasets/gradio/chatinterface_with_image_csv.clinical-narrative-image-integrity-v0.2
Clinical Narrative Image Integrity v0.2
What this is
A small dataset that tests one question:
Can you detect when a clinical narrative-image system is moving toward integrity failure, not just carrying ambiguity?
This repo focuses on narrative-image integrity under clinical reasoning pressure.
It models a system where:
narrative coherence may weaken
image alignment may drift
interpretive distortion may rise
fragmented signal may destabilize representation before overt… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-narrative-image-integrity-v0.2.ImageConfounding
Replication Data for: Integrating Earth Observation Data into Causal Inference: Challenges and Opportunities
Details:
YandW_mat.csv contains individual-level observational data. In the dataset, LONGITUDE and LATITUDE refer to the approximate geo-referenced long/lat of observational units. Experimental outcomes are stored in Yobs. The treatment variable is stored in Wobs. The unique image key for each observational unit is saved in UNIQUE_ID.
Geo-referenced satellite images are saved in… See the full description on the dataset page: https://huggingface.co/datasets/cjerzak/ImageConfounding.chatinterface_with_image_csv
Dataset Card for Dataset Name
Dataset Summary
[More Information Needed]
Supported Tasks and Leaderboards
[More Information Needed]
Languages
[More Information Needed]
Dataset Structure
Data Instances
[More Information Needed]
Data Fields
[More Information Needed]
Data Splits
[More Information Needed]
Dataset Creation
Curation Rationale
[More Information Needed]
Source Data… See the full description on the dataset page: https://huggingface.co/datasets/freddyaboulton/chatinterface_with_image_csv.imageability-corpus
English Words Imageability
This dataset is a collection of two datasets provided by Marc A. Kastner on GitHub.
I merged the datasets and kept only the word, visual, phonetic, and textual columns.
The data is scaled using a MinMaxScaler so that the whole dataset can be used as one.
Usage
This dataset is ideal for training and evaluating machine learning models for word imageability.
Acknowledgments
We extend our heartfelt gratitude to all the authors of the… See the full description on the dataset page: https://huggingface.co/datasets/StephanAkkerman/imageability-corpus.T2P
T2P: Textile-to-Physics fabric parameters
T2P is a tabular dataset of 1,382 real fabrics with physical properties expressed as
CLO3D cloth-simulation parameters, paired with each fabric's fiber composition and
construction metadata.
The task: predict a fabric's simulation-ready physical parameters (bending / shear /
stretch stiffness, buckling, friction, weight, damping) from its composition and
construction descriptors — bridging material identity ("95% cotton, 5% elastane… See the full description on the dataset page: https://huggingface.co/datasets/image2garment/T2P.socio-moral-image-rationales
Socio-Moral Image Rationales
This is a collection of machine-generated and human-labeled explanations for immorality in images.
The images are source from the Socio-Moral Image Database (SMID) and limited to the ones displaying immoral content (SMID moral mean <= 2.0).
Sampled explanations were generated by vision-language model using the ILLUME paradigm presented in ILLUME: Rationalizing Vision-Language Models through Human Interactions.
Explanations are rated by human annotators… See the full description on the dataset page: https://huggingface.co/datasets/AIML-TUDA/socio-moral-image-rationales.testgen_image_wordnet_preferencesThis dataset contains generated images. See the associated Hugging Face Collection for examples and additional details: Generated Image Wordnet
ru-image-generation
🧠 Image Generation Benchmark Dataset
This dataset simulates a large-scale benchmark for analyzing the performance of a text-to-image generation system. It contains 100,000 entries with user prompt data, generated image metadata, and multi-criteria quality ratings.
📁 Dataset Structure
Each row in the dataset corresponds to a single image generation request and includes the following fields:
Column Name
Description
request_id
Unique identifier for the request… See the full description on the dataset page: https://huggingface.co/datasets/ZennyKenny/ru-image-generation.Fitzwilliam-museum-imagesA CSV file of image urls and meta data for the Fitzwilliam Museum system. The images are licensed under more restrictive terms, the links to URLS
are open via their API and website.
ImageClassificationCatsAndDogs
ImageClassificationCatsAndDogs
tags: Classification, Object Recognition, Feline-Canine
Note: This is an AI-generated dataset so its content may be inaccurate or false
Dataset Description:
The 'ImageClassificationCatsAndDogs' dataset is designed for the purpose of image classification, specifically to distinguish between cats and dogs. Each image in the dataset is labeled with the category it belongs to, which aids in training machine learning models for object recognition. The… See the full description on the dataset page: https://huggingface.co/datasets/infinite-dataset-hub/ImageClassificationCatsAndDogs.pathorchestra-image-features
PathOrchestra Feature Representations
🔒 Access Policy
Access to this dataset is restricted and requires approval.Please request access using your official/institutional email address by contacting the dataset maintainers.
Note: Commercial use is prohibited without explicit permission.
🔄 Dataset Updates
This dataset is under continuous development as part of the broader PathOrchestra project.The current release includes the pancancer_1 subset. Additional… See the full description on the dataset page: https://huggingface.co/datasets/AI4Pathology/pathorchestra-image-features.image_caption_regularization
Regularization Image Caption Dataset
Number of Images: 1976
Source
This is a subset of tomg-group-umd/pixelprose, converted to .csv format.
Files
people.csv: 1976 images with captions that contain one of these terms: ['person', 'people', 'man', 'men', 'woman', 'women']