datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
RAM-W600
Dataset Card for RAM-W600
Benchmark code is available in https://github.com/YSongxiao/RAM-W600.
Download
Please run the following command to download RAM-W600:
git clone https://huggingface.co/datasets/TokyoTechMagicYang/RAM-W600
BoneSegmentation Mask Channel Mapping
The BoneSegmentation masks are stored as 14-channel .npy arrays with shape:
(14, H, W)
Each channel is a binary mask for one anatomical structure. The official channel order is:… See the full description on the dataset page: https://huggingface.co/datasets/TokyoTechMagicYang/RAM-W600.RAM-H1200-v1
RAM-H1200
Dataset Summary
RAM-H1200 is a multi-task full-hand radiograph dataset for rheumatoid arthritis (RA) related image analysis. It is designed to support several clinically relevant computer vision tasks, including:
hand bone structure segmentation
bone erosion related segmentation
joint localization for Sharp/van der Heijde (SvdH) scoring
joint-level SvdH bone erosion (BE) scoring
joint-level SvdH joint space narrowing (JSN) scoring
The dataset contains… See the full description on the dataset page: https://huggingface.co/datasets/TokyoTechMagicYang/RAM-H1200-v1.Sastra_ID_Cardhuashengrampnet-datasetRampNet is a two-stage pipeline that addresses the scarcity of curb ramp detection datasets by using government location data to automatically generate over 210,000 annotated Google Street View panoramas. This new dataset is then used to train a state-of-the-art curb ramp detection model that significantly outperforms previous efforts. In this repo, we provide our generated curb ramp dataset that we use to train the model.
Each parquet row contains a panoramic image and… See the full description on the dataset page: https://huggingface.co/datasets/projectsidewalk/rampnet-dataset.retina-age-analysis
Retina Age Analysis Dataset
Dataset Description
This dataset contains 9,857 retinal fundus images from 5,393 patients for age prediction tasks.
Dataset Summary
Task: Age prediction from retinal fundus images
Images: 9,857 high-quality retinal images
Patients: 5,393 unique patients
Age Range: 5-97 years
Image Format: JPEG
Average Image Size: ~1 MB
Supported Tasks
Regression: Predict continuous age (5-97 years)
Classification: Predict age group (5… See the full description on the dataset page: https://huggingface.co/datasets/ramankamran/retina-age-analysis.rampnet-crop-model-dataset-round2RampNet is a two-stage pipeline that addresses the scarcity of curb ramp detection datasets by using government location data to automatically generate over 210,000 annotated Google Street View panoramas. This new dataset is then used to train a state-of-the-art curb ramp detection model that significantly outperforms previous efforts. In this repo, we provide "the tiny set of manually labeled crops" that we refer to in both RampNet's GitHub repository and the paper.
Renamed 2026-08-05: this… See the full description on the dataset page: https://huggingface.co/datasets/projectsidewalk/rampnet-crop-model-dataset-round2.ramanv-image-editing
ramanv-image-editing
Image editing dataset for training FLUX.1-Kontext / InstructPix2Pix style models.
Size
592,141 total editing pairs
Sources: ultraedit
Schema
Each shard tar contains {uid}_src.jpg, {uid}_edit.jpg, {uid}_mask.png (where available).
Metadata per record: instruction, prompt, edit_type, caption_before/after, license, sha256.
Licenses
MagicBrush, InstructPix2Pix, Pico-Banana, HumanEdit: CC-BY-4.0
UltraEdit, AnyEdit… See the full description on the dataset page: https://huggingface.co/datasets/lingamvamshikrishnareddy/ramanv-image-editing.ramenakaneko
Bangumi Image Base of Ramen Akaneko
This is the image base of bangumi Ramen Akaneko, we detected 33 characters, 1274 images in total. The full dataset is here.
Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual. If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters'… See the full description on the dataset page: https://huggingface.co/datasets/BangumiBase/ramenakaneko.SynthCheX-75K-v2
SynthCheX-75K
SynthCheX-75K is released as a part of the CheXGenBench paper. It is a synthetic dataset generated using Sana (0.6B) [1] fine-tuned on chest radiographs. Sana (0.6B) establishes the SoTA performance on the CheXGenBench benchmark.
The dataset contains 75,649 high-quality image-text samples along with the pathological annotations.
Filtration Process for SynthCheX-75K
Generative models can lead to both high and low-fidelity generations on different subsets… See the full description on the dataset page: https://huggingface.co/datasets/raman07/SynthCheX-75K-v2.ramen_benchmark_jp_beirThis is a copy of https://huggingface.co/datasets/jinaai/ramen_benchmark_jp reformatted into the BEIR format. For any further information like license, please refer to the original dataset.
Disclaimer
This dataset may contain publicly available images or text data. All data is provided for research and educational purposes only. If you are the rights holder of any content and have concerns regarding intellectual property or copyright, please contact us at "support-data (at) jina.ai"… See the full description on the dataset page: https://huggingface.co/datasets/jinaai/ramen_benchmark_jp_beir.sim_ram_scriptedThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.0",
"robot_type": "aloha",
"total_episodes": 150,
"total_frames": 60000,
"total_tasks": 1,
"total_videos": 0,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 50,
"splits": {
"train": "0:150"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/hongdaaaaaaaa/sim_ram_scripted.sim_ram_scripted_2This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.0",
"robot_type": "aloha",
"total_episodes": 100,
"total_frames": 40000,
"total_tasks": 1,
"total_videos": 0,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 50,
"splits": {
"train": "0:100"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/hongdaaaaaaaa/sim_ram_scripted_2.plant-diseases-100kDiabetic_Retinopathy_Preprocessed_Dataset_256x256This is dataset comes from this Kaggle Dataset
from the user Sachin Kumar.
The goal of the dataset is for the Varun AIM Projects to easily start running and download the dataset on their local computer in the HF libraries as the directory I strongly recommedn to use.
automatum-data-highway-with-ramps
Automatum Data: Highway with Ramps Drone Dataset
Introduction
The Automatum Data Highway with Ramps Dataset contains high-precision movement data of traffic participants (cars, trucks, vans) extracted from drone recordings on German Autobahn segments with on-ramps and off-ramps. Captured from a bird's eye view, the dataset provides complete trajectories with velocities, accelerations, lane assignments, and object relationships — perfectly suited for merging… See the full description on the dataset page: https://huggingface.co/datasets/AutomatumData/automatum-data-highway-with-ramps.SidewalkPilot_v1_and_v2
SidewalkPilot Series 1 and 2 Steering Dataset
SidewalkPilot Series 1 and 2 is the finalized camera-to-steering dataset for the baseline and failure/iteration model series. The dataset pairs real field images with steering servo labels in degrees, so a model can learn to map a camera frame to a steering command.
CARLA-assisted. The Series 1/2 models were trained on a blend of these real field images plus CARLA synthetic frames (down-weighted vs real). This repository holds the… See the full description on the dataset page: https://huggingface.co/datasets/ram-shreyas-naik-sabavat/SidewalkPilot_v1_and_v2.indic_wikisourceDataset contains page image urls and the corresponding annotations from wikisource.
Also has information whether the page has been validated/proofread.
I expect this dataset to be useful for creating OCR models for printed text in Indic languages.
Languages Covered:
as - Assamese
bn - Bengali
gu - Gujarati
hi - Hindi
kn - Kannada
ml - Malayalam
mr - Marathi
or - Odiya
pa - Punjabi
sa - Sanskrit
ta - Tamil
te - Telugu
ramanv-domain-trainingsim_ram_scripted_3This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.0",
"robot_type": "aloha",
"total_episodes": 100,
"total_frames": 40000,
"total_tasks": 1,
"total_videos": 0,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 50,
"splits": {
"train": "0:100"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/hongdaaaaaaaa/sim_ram_scripted_3.test-flesrampnet-benchmark
RampNet Benchmark Imagery
⚠️ This benchmark is not part of the RampNet paper
It did not exist when RampNet was published. The paper's tag,
v1.0-iccv2025 (August 2025),
contains no benchmark/ directory at all — its evaluation was a 1,000-panorama manually
labeled gold set (manual_labels/, imagery in
rampnet-dataset), drawn from
the same three training cities.
These 9 city splits were built eleven months later, between 2026-07-22 and 2026-08-01, as
post-publication… See the full description on the dataset page: https://huggingface.co/datasets/projectsidewalk/rampnet-benchmark.rampnet-crop-model-dataset-round1
RampNet Crop-Model Dataset — Round 1 (Project Sidewalk crops)
The training data behind round 1 of the RampNet Stage 1 crop model — 27,704 crops,
13.37 GB — from RampNet: A Two-Stage Pipeline for Bootstrapping Curb Ramp Detection in
Streetscape Images from Open Government Metadata (O'Meara et al., ICCV'25 CV4A11y workshop,
arXiv:2508.09415).
The crop model is what turns a government curb ramp GPS coordinate into a pixel keypoint on a
panorama; every label in
rampnet-dataset was… See the full description on the dataset page: https://huggingface.co/datasets/projectsidewalk/rampnet-crop-model-dataset-round1.food-items-indonesianramanv-image-rawphone_teleop_testiphone_stairs_ramps
iphone_stairs_ramps
Description
Processed entire iphone_stairs_ramps with filter_every_nth=1, 100% of data, and num_subsampled_points=5
Processing Parameters
mateoguaman/iphone_chin:
exclude_outliers_pct: 0
filter_by_curvature: false
filter_every_nth: 1
horizon:
1000: 1.0
num_subsampled_points: 5
mateoguaman/iphone_hip:
exclude_outliers_pct: 0
filter_by_curvature: false
filter_every_nth: 1
horizon:1000: 1.0
num_subsampled_points: 5… See the full description on the dataset page: https://huggingface.co/datasets/mateoguaman/iphone_stairs_ramps.dm_lab
UNDER CONSTRUCTION !!!
DeepMind-Lab 30 Benchmark
This dataset contains expert trajectories generated by a Dreamer V3 reinforcement learning agent trained on each of the 30 environments in DMLab-30. Contrary to other datasets, we provide image observations instead of states.
Dataset Usage
Regular usage (for the domain acrobot with task swingup):
from datasets import load_dataset
train_dataset = load_dataset("EpicPinkPenguin/visual_dm_control"… See the full description on the dataset page: https://huggingface.co/datasets/ramu0e/dm_lab.smallnorb
Dataset Card for "smallnorb"
Dataset Description
NOTE: This dataset is an unofficial port of small NORB based on a repo from Andrea Palazzi using this script. For complete and accurate information, we highly recommend visiting the dataset's original homepage.
Homepage: https://cs.nyu.edu/~ylclab/data/norb-v1.0-small/
Paper: https://ieeexplore.ieee.org/document/1315150
Dataset Summary
From the dataset's homepage:
This database is intended for experiments in… See the full description on the dataset page: https://huggingface.co/datasets/Ramos-Ramos/smallnorb.audiofeaturesalbumcovers
Dataset Card for "audiofeaturesalbumcovers"
More Information needed