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.mask-for-image-segmentation-testsdroid_dataset_segmentation_mask
DROID SAM 3.1 Segmentation Masks
This dataset is a mask-only sidecar generated from the original
droid_101/0.0.1 RLDS release. It does not redistribute DROID images or
actions. Its episode_index follows the RLDS episode order.
The same episodes appear in lerobot/droid_1.0.1, but LeRobot stores them in a
different episode order. Therefore, mask episode_index and LeRobot
episode_index must not be joined directly. Use the mapping file described
below to associate these masks with… See the full description on the dataset page: https://huggingface.co/datasets/EpicPinkPenguin/droid_dataset_segmentation_mask.053025-segmentation-masksLeaf-Disease-Segmentation-masks
Leaf-Disease-Segmentation-masks
Dataset Description
This dataset contains in-the-wild images of plant leaves and their corresponding binary segmentation masks. It is intended for training computer vision models for segmentation tasks.
Features
image: The original RGB image of the plant leaf/leaves.
mask: The binary, single-channel (grayscale) segmentation mask.
Dataset Structure
The dataset consists of 588 image-mask pairs.
segmentation-masksadaptive-segmentation-mask-attackrandall-segmentation-masks
Randall Manuscript Segmentation Masks
This dataset contains all hand-annotated segmentation masks for manuscripts indexed in
Lilian Randall's Images in the Margins of Gothic Manuscripts (1966).
It contains 482 training masks and 154 test masks across 87 manuscript groups.
7 categories of decorative features are labeled. Text, blank page, and other non-decorative
regions are grouped together as background.
Fields
name, uuid, labels, segments_metadata, and… See the full description on the dataset page: https://huggingface.co/datasets/wellesley-easel/randall-segmentation-masks.hocap-segmentation-maskscmu_panoptic_dataset_instance_segmentation_mask
CMU Panoptic person instance masks (SAM 3.1 pseudo-labels)
TODO before making this repo public: confirm permission from the CMU Panoptic
Studio organisers. Until then the repo must stay gated and private. See
LICENSE_NOTICE.md.
Per-person instance segmentation masks for the CMU Panoptic Studio dataset, generated
with SAM 3.1 (facebook/sam3.1 @ daa63191845a41281374e725f4c9e51c7a824460) and aligned to the
dataset's own 3D skeletons: masks and pose share the same person id across… See the full description on the dataset page: https://huggingface.co/datasets/EpicPinkPenguin/cmu_panoptic_dataset_instance_segmentation_mask.luna16-10-percent-segmentation-maskssegmentation-masks-dataset