datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
pixmo-points
PixMo-Points
PixMo-Points is a dataset of images paired with referring expressions and points marking the locations the
referring expression refers to in the image. It was collected using human annotators and contains a diverse
range of points and expressions, with many high-frequency (10+) expressions.
PixMo-Points is a part of the PixMo dataset collection and was used to
provide the pointing capabilities of the Molmo family of models
Quick links:
📃 Paper
🎥 Blog with Videos… See the full description on the dataset page: https://huggingface.co/datasets/allenai/pixmo-points.ro_sft_pixmo_points
Dataset Description
PixmoPoints is a dataset of images paired with referring expressions and points marking the locations the referring expression refers to in the image.
Here we provide the Romanian translation of the PixmoPoints dataset, translated with Seed-X-PPO.
This dataset is part of the instruction finetune protocol for Romanian VLMs proposed in "Înțelegi românește?" A Recipe for Romanian Vision-Language Models (Masala et al., 2026).
Citation… See the full description on the dataset page: https://huggingface.co/datasets/surogate/ro_sft_pixmo_points.xarm_dishwasher_points9_arrow_len0
xarm_dishwasher_points9_arrow_len0
xArm7 phone-teleop dataset for dishwasher ("pull the basket outside of the dishwasher and pick up the mug and put it into the basket") with the points9_arrow_len0
tactile overlay burned into the camera frames: 9 dots per gripper finger drawn at their
real 3D tactile-pad positions (forward kinematics, projected through camera intrinsics),
with force arrows rendered at zero length (arrow_length_scale=0).
This is the force-ablation control for the… See the full description on the dataset page: https://huggingface.co/datasets/EdwardoSunny/xarm_dishwasher_points9_arrow_len0.pixmo-pointsxarm_charger_points9_arrow_len0
xarm_charger_points9_arrow_len0
xArm7 phone-teleop dataset for charger ("pick up the charger and plug it into the nearest plug") with the points9_arrow_len0
tactile overlay burned into the camera frames: 9 dots per gripper finger drawn at their
real 3D tactile-pad positions (forward kinematics, projected through camera intrinsics),
with force arrows rendered at zero length (arrow_length_scale=0).
This is the force-ablation control for the points9_arrow variant: identical dot… See the full description on the dataset page: https://huggingface.co/datasets/EdwardoSunny/xarm_charger_points9_arrow_len0.pixmo-points-filtered-below10_imgContainedpixmo-points-evalpixmo-points-eval
PixMo-Points-Eval
PixMo-Points-Eval is a subset of PixMo-Points that has been human-filtered and annotated with segmentation masks.
It is used for pointing evaluations.
PixMo-Points is a part of the PixMo dataset collection and was used to
provide the pointing capabilities of the Molmo family of models
Loading
data = datasets.load_dataset("allenai/pixmo-points-eval", split="test")
Data Format
Images are stored as URLs that will need to be downloaded… See the full description on the dataset page: https://huggingface.co/datasets/allenai/pixmo-points-eval.xarm_cube_points9_arrow_len0
xarm_cube_points9_arrow_len0
xArm7 phone-teleop dataset for cube ("pick up the cube") with the points9_arrow_len0
tactile overlay burned into the camera frames: 9 dots per gripper finger drawn at their
real 3D tactile-pad positions (forward kinematics, projected through camera intrinsics),
with force arrows rendered at zero length (arrow_length_scale=0).
This is the force-ablation control for the points9_arrow variant: identical dot
positions and visual salience, but zero force… See the full description on the dataset page: https://huggingface.co/datasets/EdwardoSunny/xarm_cube_points9_arrow_len0.PointSplat_example_data
PointSplat Example Data
This repository hosts compact example data for running PointSplat inference locally.
The zip archives are the recommended download format for reproducing the commands in the PointSplat README. The Dataset Viewer shows a small preview split with one fixed camera per scene:
DNA-Rendering: 16 validation scenes, camera 23, frame 000030.
THuman2.0: 10 validation scenes, view 000, frame 2.
Files
File
Content
Local target path… See the full description on the dataset page: https://huggingface.co/datasets/Yujie0012/PointSplat_example_data.xarm_tube_points9_arrow_len0
xarm_tube_points9_arrow_len0
xArm7 phone-teleop dataset for tube ("pick up the tube and put it in the nearest slot") with the points9_arrow_len0
tactile overlay burned into the camera frames: 9 dots per gripper finger drawn at their
real 3D tactile-pad positions (forward kinematics, projected through camera intrinsics),
with force arrows rendered at zero length (arrow_length_scale=0).
This is the force-ablation control for the points9_arrow variant: identical dot
positions and… See the full description on the dataset page: https://huggingface.co/datasets/EdwardoSunny/xarm_tube_points9_arrow_len0.Generation-of-Virtual-Ground-Control-Points-using-a-Binocular-Camera-Datasetsro_sft_pixmo_points
Dataset Description
PixmoPoints is a dataset of images paired with referring expressions and points marking the locations the referring expression refers to in the image.
Here we provide the Romanian translation of the PixmoPoints dataset, translated with Seed-X-PPO.
This dataset is part of the instruction finetune protocol for Romanian VLMs proposed in "Înțelegi românește?" A Recipe for Romanian Vision-Language Models (Masala et al., 2026).
Citation… See the full description on the dataset page: https://huggingface.co/datasets/OpenLLM-Ro/ro_sft_pixmo_points.pixmo-points
PixMo-Points
PixMo-Points is a dataset of images paired with referring expressions and points marking the locations the
referring expression refers to in the image. It was collected using human annotators and contains a diverse
range of points and expressions, with many high-frequency (10+) expressions.
PixMo-Points is a part of the PixMo dataset collection and was used to
provide the pointing capabilities of the Molmo family of models
Quick links:
📃 Paper
🎥 Blog with Videos… See the full description on the dataset page: https://huggingface.co/datasets/mlmPenguin/pixmo-points.ro_pixmo_points_eval
Dataset Description
Pixmo-Points is a dataset that has been human-filtered and annotated with segmentation masks. It is used for pointing evaluations.
Here we provide the Romanian translation of Pixmo-Points test set, translated with gpt-4.1-mini. This dataset is used as a benchmark and is part of the evaluation protocol for Romanian VLMs proposed in "Înțelegi românește?" A Recipe for Romanian Vision-Language Models (Masala et al., 2026).
Citation… See the full description on the dataset page: https://huggingface.co/datasets/OpenLLM-Ro/ro_pixmo_points_eval.kvasir-points-v1easyr1-10k-hard-qwen7b-easy-gta1-4MP-no-pixmo-points
easyr1-10k-hard-qwen7b-easy-gta1-4MP-no-pixmo-points
This dataset was generated using the EasyR1 grounding dataset pipeline.
Generation Details
Generated on: 2025-08-23 16:05:22 UTC
Script: push_easyr1_to_hf.py
Data directory: /lustre/fsw/portfolios/nvr/users/aawadalla/LLaMA-Factory/data
Parameters Used
Maximum samples: 10000
Image resize (max megapixels): 4.0 MP
Minimum native image resolution: 0.0 MP
Prompt format: gta1_with_resolution
Output format:… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations-cua-dev/easyr1-10k-hard-qwen7b-easy-gta1-4MP-no-pixmo-points.pixmo-points-filtered-below10pixmo-points-filtered_0-20_imgContainedpixmo_points_dataset_50000Pixmo-Points-RL-8Keasyr1-pixmo-points-only-4k9-omniparser-qwen-tool-call-4MP
easyr1-pixmo-points-only-4k9-omniparser-qwen-tool-call-4MP
This dataset was generated using the EasyR1 grounding dataset pipeline.
Generation Details
Generated on: 2025-08-21 12:54:12 UTC
Script: push_easyr1_to_hf.py
Data directory: /lustre/fsw/portfolios/nvr/users/aawadalla/LLaMA-Factory/data
Parameters Used
Maximum samples: 4888
Image resize (max megapixels): 4.0 MP
Prompt format: gta1_with_resolution
Output format: coordinates
Random seed: 42
Resampling… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations-cua-dev/easyr1-pixmo-points-only-4k9-omniparser-qwen-tool-call-4MP.ena24-points-test
ena24-points-test
Auto-labeled with MolmoPoint using the query: "point to the animals"
Source
Input dataset: davanstrien/ena24-detection
Model: allenai/MolmoPoint-8B
Stats
Metric
Value
Images processed
5
Images with detections
0 (0.0%)
Total points detected
0
Avg points/image
0.0
Output columns
points_x: List of x-coordinates (pixels) for detected objects
points_y: List of y-coordinates (pixels) for detected objects… See the full description on the dataset page: https://huggingface.co/datasets/davanstrien/ena24-points-test.ena24-points-test2
ena24-points-test2
Auto-labeled with MolmoPoint using the query: "point to the animals"
Source
Input dataset: davanstrien/ena24-detection
Model: allenai/MolmoPoint-8B
Stats
Metric
Value
Images processed
5
Images with detections
2 (40.0%)
Total points detected
2
Avg points/image
0.4
Output columns
points_x: List of x-coordinates (pixels) for detected objects
points_y: List of y-coordinates (pixels) for detected objects… See the full description on the dataset page: https://huggingface.co/datasets/davanstrien/ena24-points-test2.pixmo_points_dataset_ALL_valid_imagespointspixmo_points_dataset_ALL_which_points_less_than_100pixmo_points_dataset_300pixmo_points_dataset_url_downloaded_but_maybe_htmlpixmo_points_dataset_ALL_with_smoke_fire_which_points_less_than_100