datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Pleural-Line-Segmentation-Masks
Pleural-Line Masks with Stanford LUS Frames
Dataset Summary
This dataset contains pleural-line masks and their corresponding lung-ultrasound frames for anatomy-guided video classification.
The mask set includes:
masks created by four human annotators via sam2 model point promting and video aggregation;
masks predicted by a U-Net and subsequently reviewed and validated; and
the metadata required to reproduce the training pipeline.
The ultrasound frames originate… See the full description on the dataset page: https://huggingface.co/datasets/alyaalmsouti/Pleural-Line-Segmentation-Masks.BraTS_segmentationimage-segmentation-toy-datauniversal-lesion-segmentation
Universal Lesion Segmentation Datasets
A collection of public medical imaging datasets for lesion segmentation in CT scans. These are the datasets exactly as downloaded from their original sources.
Datasets
This repository contains the following datasets:
CECT - Liver (primary). Luo J, Wang X, Zhang Y, et al. Comprehensive multi-phase three-dimensional contrast-enhanced CT imaging dataset for primary liver cancer. Scientific Data. 2025;12(1):768.… See the full description on the dataset page: https://huggingface.co/datasets/nielsRocholl/universal-lesion-segmentation.lgg-mri-segmentation-research
LGG Brain MRI Segmentation with Genomic Clusters
This repository provides a Patient-Centric version of the Lower-Grade Glioma (LGG) Segmentation dataset. While other versions of this data exist, they often treat slices as independent images. This version preserves the 3D patient volume and integrates all genomic/clinical labels directly into a multimodal-ready format.
🌟 Why This Version?
Developed for Multimodal AI Research, this dataset addresses several limitations… See the full description on the dataset page: https://huggingface.co/datasets/Ehsan-rmz/lgg-mri-segmentation-research.Imbalanced_Segmentation_DatasetsBDD100K-Semantic-SegmentationBDD100K: Semantic Segmentation Dataset
Unofficial redistribution of BDD100K's semantic segmentation task (10K-image subset), under BDD100K's own data license (non-commercial/educational/research redistribution, with notice, explicitly permitted).
Disclaimer
This repository is not an official release of BDD100K.
BDD100K was created by Fisher Yu, Haofeng Chen, Xin Wang, Wenqi Xian, Yingying Chen, Fangchen Liu, Vashisht Madhavan, and Trevor Darrell at UC Berkeley… See the full description on the dataset page: https://huggingface.co/datasets/dronefreak/BDD100K-Semantic-Segmentation.table_spill_cleanup_bimanual_rgbd_segmentation_poses
Exylos Bimanual Table Spill Cleanup Rich-Modality Sample
A compact, rich-modality bimanual robot manipulation dataset for tabletop spill cleanup.
Each episode combines synchronized dual-arm Panda state/action trajectories, 7 RGB camera streams, per-frame depth maps, per-frame segmentation masks, object pose streams, phase annotations, and an objective cleanup success metric based on the remaining spill fraction.
This dataset is a rich-modality inspection sample for the Exylos… See the full description on the dataset page: https://huggingface.co/datasets/ExylosAi/table_spill_cleanup_bimanual_rgbd_segmentation_poses.wmh-segmentation
WMH Segmentation Challenge Dataset
Dataset Description
The WMH Segmentation Challenge dataset for white matter hyperintensities segmentation. This dataset contains MRI FLAIR scans with dense segmentation annotations.
Dataset Details
Modality: MRI FLAIR
Target: white matter hyperintensities
Format: NIfTI (.nii.gz)
Dataset Structure
Each sample in the JSONL file contains:
{
"image": "path/to/image.nii.gz",
"mask":… See the full description on the dataset page: https://huggingface.co/datasets/MedOtter/wmh-segmentation.SpectralWaste-Segmentation
SpectralWaste Segmentation → FiftyOne (Grouped RGB + Hyperspectral)
The labeled split of SpectralWaste, rebuilt as a grouped FiftyOne dataset pairing each colour frame with its hyperspectral cube.
The recordings come from a working waste-sorting plant, looking down at the conveyor as material passes. Each frame is captured twice over, once in colour and once by a shortwave infrared camera reading 224 bands from about 900 to 1700 nm. Material that looks identical in colour… See the full description on the dataset page: https://huggingface.co/datasets/Voxel51/SpectralWaste-Segmentation.Medical_Segmentation_Decathlon
🏆 Medical Segmentation Decathlon Dataset
📝 Overview
The Medical Segmentation Decathlon (MSD) is a comprehensive benchmark dataset for validating algorithms in 3D medical image segmentation. It includes 10 distinct tasks, each with unique challenges like small data sizes, unbalanced labels, varying object scales, multi-class labels, and multimodal imaging.
🔗 Dataset Access
🌐 Website: Medical Decathlon
📂 Google Drive: MSD Google Drive
🧩 Task… See the full description on the dataset page: https://huggingface.co/datasets/Novel-BioMedAI/Medical_Segmentation_Decathlon.Football-Player-Segmentation
Dataset Card for football-player-segmentation
This dataset is specifically designed for computer vision tasks related to player detection and segmentation in foot goalkeeperders, and forwards, captured from various angles and distances.
This is a FiftyOne dataset with 512 samples.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
import fiftyone.utils.huggingface as fouh
# Load the dataset
#… See the full description on the dataset page: https://huggingface.co/datasets/Voxel51/Football-Player-Segmentation.live-cell-segmentation-dataset
Light-microscopy cell segmentation dataset
Transmitted-light fields -- phase contrast, brightfield, DIC and quantitative phase -- with
uint16 instance masks, from 14 public datasets, curated for spaCR and used to fine-tune
live-cell-segmentation-cpsam.
11,007 fields: 6,778 train, 2,030 valid,
2,199 test. Split by ACQUISITION -- a well, a dish, a time-lapse or a
z-stack is never divided between sets -- so the test set is fields the model never saw
anything of.
Layout… See the full description on the dataset page: https://huggingface.co/datasets/einarolafsson/live-cell-segmentation-dataset.minne-apple-segmentationA version of MinneApple with coco annotations
AVM_Segmentation_train
Dataset Card for AVM (Around View Monitoring) Semantic Segmentation Dataset
This repository provides a FiftyOne-compatible version of the AVM semantic segmentation dataset for autonomous parking systems, with enhanced metadata and visualization capabilities.
This is a FiftyOne dataset with 6763 samples.
Installation
If you haven't already, install FiftyOne:
pip install -U fiftyone
Usage
import fiftyone as fo
from fiftyone.utils.huggingface import… See the full description on the dataset page: https://huggingface.co/datasets/Voxel51/AVM_Segmentation_train.mask-for-image-segmentation-testsESK_action_segmentationPaper | GitHub
🍳 EPFL-Smart-Kitchen: Action segmentation benchmark
📚 Introduction
Given an untrimmed video, action segmentation requires the model to predict one or multiple action classes for every frame.Given the absence of popular (and comprehensive) action segmentation benchmarks from 3D pose we built an action segmentation benchmark that compares the impact of different input data (body,hand,eyes, videofeatures).One might expect that actions such as moving through… See the full description on the dataset page: https://huggingface.co/datasets/amathislab/ESK_action_segmentation.robotic-instrument-segmentation-miccai-2017fire-segmentation-dataset
Fire Segmentation Dataset (YOLO-seg format)
Instance-segmentation dataset for fire detection: 1,348 images with
polygon mask labels in Ultralytics YOLO segmentation format. Built to train
sreeharivp23/fire-segmentation-yolo11n.
Contents
Split
Images
train
1,146
val
202
1,098 fire images with one or more fire polygon instances
250 negatives (no fire) with empty label files
fire_seg/
├── data.yaml # Ultralytics dataset config (1 class:… See the full description on the dataset page: https://huggingface.co/datasets/sreeharivp23/fire-segmentation-dataset.news-segmentationROI-1555_Rebar_Detection_and_Instance_Segmentation_DatasetROI-1555: Rebar Detection and Instance Segmentation Dataset
ROI-1555 for rebar object detection and instance segmentation contains 1555 rebar images and their fine-labeled bounding boxes and pixel-wise masks.
Diverse rebar specifications, layouts, application scenarios, and environmental conditions.
Usage
Here is an example to convert the annotations to MSCOCO 2017 format
python
cp -r 1260/img_label tools/data_annotated/train2017
cd tools
python labelme2coco_instance.py… See the full description on the dataset page: https://huggingface.co/datasets/tsrobcvai/ROI-1555_Rebar_Detection_and_Instance_Segmentation_Dataset.tunnel-crack-segmentation-dataset
Complex Tunnel Crack Dataset (CTCD)
292 tunnel lining images (256 × 256) with pixel-level crack masks, collected from
52 highway tunnels in Yunnan and Zhejiang, China, using a high-speed tunnel
inspection vehicle. Crack pixels were manually
annotated in Photoshop with a 1–2 px brush sized by crack width.
This dataset was introduced in the CraSAM paper (Structural Control and Health
Monitoring, 2025), where it is used to evaluate cross-scenario generalization of… See the full description on the dataset page: https://huggingface.co/datasets/shiweiluo99/tunnel-crack-segmentation-dataset.vhr-building-segmentation
HOT Building Segmentation Dataset
Dataset Description
A semantic segmentation dataset for building footprint extraction from aerial imagery, built from validated Humanitarian OpenStreetMap Team (HOT) Tasking Manager projects that use OpenAerialMap (OAM) imagery.
Dataset Summary
This dataset pairs 256x256 aerial image tiles (zoom level 19) from OpenAerialMap with building footprint labels from OpenStreetMap. All source projects have been fully… See the full description on the dataset page: https://huggingface.co/datasets/hotosm/vhr-building-segmentation.object-segmentationRS_Image_Segmentation_Vaihingenpothole-segmentation
Dataset Labels
['pothole']
Number of Images
{'test': 5, 'train': 80, 'valid': 5}
How to Use
Install datasets:
pip install datasets
Load the dataset:
from datasets import load_dataset
ds = load_dataset("keremberke/pothole-segmentation", name="full")
example = ds['train'][0]
Roboflow Dataset Page
https://universe.roboflow.com/imacs-pothole-detection-wo8mu/pothole-detection-irkz9/dataset/4
Citation
@misc{… See the full description on the dataset page: https://huggingface.co/datasets/keremberke/pothole-segmentation.semantic-segmentation-test-sampleThis dataset contains 10 examples of the segments/sidewalk-semantic dataset (i.e. 10 images with corresponding ground-truth segmentation maps).
segmentation-carla-drivingThis dataset consists of 80 episodes of driving data collected using an autopilot agent in CARLA simulator for training imitation learning models for autonomous driving tasks.
Each frame is structured as follows:
frame_data = {
'frame': the frame index,
'hlc': an integer representing the high-level command,
'light': an integer representing current traffic light status,
'controls': an array of [throttle, steer, brake],
'measurements':… See the full description on the dataset page: https://huggingface.co/datasets/nightmare-nectarine/segmentation-carla-driving.robotic-scene-segmentation-miccai-2018pcb-defect-segmentation
Dataset Labels
['dry_joint', 'incorrect_installation', 'pcb_damage', 'short_circuit']
Number of Images
{'valid': 25, 'train': 128, 'test': 36}
How to Use
Install datasets:
pip install datasets
Load the dataset:
from datasets import load_dataset
ds = load_dataset("keremberke/pcb-defect-segmentation", name="full")
example = ds['train'][0]
Roboflow Dataset Page
https://universe.roboflow.com/diplom-qz7q6/defects-2q87r/dataset/8… See the full description on the dataset page: https://huggingface.co/datasets/keremberke/pcb-defect-segmentation.