datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
reef-guidance-system
Dataset Card for Reef Guidance System
This dataset provides imagery used for training and evaluation of models in the Reef Guidance System. All imagery was collected by the Australian Institute of Marine Science using the ReefScan™ Transom Marine Monitoring System.
If you use this dataset in your work, please cite the associated paper: AI-driven dispensing of coral reseeding devices for broad-scale restoration of the Great Barrier Reef (citations provided at bottom of this… See the full description on the dataset page: https://huggingface.co/datasets/QCR-Underwater-Perception/reef-guidance-system.gently-perception-benchmark
Gently Perception Agent Benchmark
Light-sheet microscopy volumes of C. elegans embryo development, intended
for evaluating vision-based perception agents on embryo stage classification.
The dataset has two tiers:
Annotated benchmark set (embryo_1–embryo_8) — human ground-truth
stage transitions. Use this for evaluation.
Unannotated corpus (embryo_9–embryo_105) — 97 additional real embryo
timelapses with no human labels, provided for developing and stress-
testing perception… See the full description on the dataset page: https://huggingface.co/datasets/gently-project/gently-perception-benchmark.lighting-invariant-bedroom-perception-robustness-benchmark
Lighting-Invariant Bedroom Perception & Robustness Benchmark
Generated by datapack-import.ts
This dataset mirrors public data-pack render outputs from Physicl.
Each row represents one render view. The image column contains a stable URL to the primary render image uploaded under /data; image_path stores the relative repository path and data_commit_sha pins the Hugging Face dataset commit used by those URLs. Files are uploaded as downloaded unless optional PNG recompression is… See the full description on the dataset page: https://huggingface.co/datasets/physicl/lighting-invariant-bedroom-perception-robustness-benchmark.synthetic-bathroom-dataset-for-robotic-perception
Synthetic Bathroom Dataset for Robotic Perception
Generated by datapack-import.ts
This dataset mirrors public data-pack render outputs from Physicl.
Each row represents one render view. The image column contains a stable URL to the primary render image uploaded under /data; image_path stores the relative repository path and data_commit_sha pins the Hugging Face dataset commit used by those URLs. Files are uploaded as downloaded unless optional PNG recompression is enabled by the… See the full description on the dataset page: https://huggingface.co/datasets/physicl/synthetic-bathroom-dataset-for-robotic-perception.imaginative-perception-token-mvc-ipt
Citation
Released with the paper Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models (arXiv:2606.03988):
@misc{bigverdi2026imaginativeperceptiontokensenhance,
title={Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models},
author={Mahtab Bigverdi and Linjie Li and Weikai Huang and Yiming Liu and Jaemin Cho and Jieyu Zhang and Tuhin Kundu and Chris Dangjoo Kim and Zelun Luo and Linda Shapiro and Ranjay… See the full description on the dataset page: https://huggingface.co/datasets/weikaih/imaginative-perception-token-mvc-ipt.AirV2X-Perceptionimaginative-perception-token-pet-ipt
Citation
Released with the paper Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models (arXiv:2606.03988):
@misc{bigverdi2026imaginativeperceptiontokensenhance,
title={Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models},
author={Mahtab Bigverdi and Linjie Li and Weikai Huang and Yiming Liu and Jaemin Cho and Jieyu Zhang and Tuhin Kundu and Chris Dangjoo Kim and Zelun Luo and Linda Shapiro and Ranjay… See the full description on the dataset page: https://huggingface.co/datasets/weikaih/imaginative-perception-token-pet-ipt.synthetic-living-room-dataset-for-robotic-perception
Synthetic Living Room Dataset for Robotic Perception
Generated by datapack-import.ts
This dataset mirrors public data-pack render outputs from Physicl.
Each row represents one render view. The image column contains a stable URL to the primary render image uploaded under /data; image_path stores the relative repository path and data_commit_sha pins the Hugging Face dataset commit used by those URLs. Files are uploaded as downloaded unless optional PNG recompression is enabled by… See the full description on the dataset page: https://huggingface.co/datasets/physicl/synthetic-living-room-dataset-for-robotic-perception.imaginative-perception-token-pet-eval-ai2thor
Citation
Released with the paper Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models (arXiv:2606.03988):
@misc{bigverdi2026imaginativeperceptiontokensenhance,
title={Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models},
author={Mahtab Bigverdi and Linjie Li and Weikai Huang and Yiming Liu and Jaemin Cho and Jieyu Zhang and Tuhin Kundu and Chris Dangjoo Kim and Zelun Luo and Linda Shapiro and Ranjay… See the full description on the dataset page: https://huggingface.co/datasets/weikaih/imaginative-perception-token-pet-eval-ai2thor.imaginative-perception-token-pet-eval-habitat
Dataset Card for "habitat_perspective_eval"
More Information needed
Citation
Released with the paper Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models (arXiv:2606.03988):
@misc{bigverdi2026imaginativeperceptiontokensenhance,
title={Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models},
author={Mahtab Bigverdi and Linjie Li and Weikai Huang and Yiming Liu and Jaemin Cho and Jieyu Zhang… See the full description on the dataset page: https://huggingface.co/datasets/weikaih/imaginative-perception-token-pet-eval-habitat.imaginative-perception-token-mvc-eval-ai2thor
Dataset Card for "multiview_eval"
More Information needed
Citation
Released with the paper Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models (arXiv:2606.03988):
@misc{bigverdi2026imaginativeperceptiontokensenhance,
title={Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models},
author={Mahtab Bigverdi and Linjie Li and Weikai Huang and Yiming Liu and Jaemin Cho and Jieyu Zhang and Tuhin… See the full description on the dataset page: https://huggingface.co/datasets/weikaih/imaginative-perception-token-mvc-eval-ai2thor.imaginative-perception-token-pt-eval-ai2thor
Citation
Released with the paper Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models (arXiv:2606.03988):
@misc{bigverdi2026imaginativeperceptiontokensenhance,
title={Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models},
author={Mahtab Bigverdi and Linjie Li and Weikai Huang and Yiming Liu and Jaemin Cho and Jieyu Zhang and Tuhin Kundu and Chris Dangjoo Kim and Zelun Luo and Linda Shapiro and Ranjay… See the full description on the dataset page: https://huggingface.co/datasets/weikaih/imaginative-perception-token-pt-eval-ai2thor.perception_lm_test_imagesvideo-diffusion-perceptionimaginative-perception-token-pt-eval-real
Citation
Released with the paper Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models (arXiv:2606.03988):
@misc{bigverdi2026imaginativeperceptiontokensenhance,
title={Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models},
author={Mahtab Bigverdi and Linjie Li and Weikai Huang and Yiming Liu and Jaemin Cho and Jieyu Zhang and Tuhin Kundu and Chris Dangjoo Kim and Zelun Luo and Linda Shapiro and Ranjay… See the full description on the dataset page: https://huggingface.co/datasets/weikaih/imaginative-perception-token-pt-eval-real.perception-benchmarkgreenery-perception
It is not always greener on the other side: Greenery perception across demographics and personalities in multiple cities
Repository for additional data used in the paper "It is not always greener on the other side: Greenery perception across demographics and personalities in multiple cities".
This project was developed at the Future Cities Lab Global in the Singapore-ETH Centre in close collaboration with the Urban Analytics Lab (UAL) at the National University of Singapore (NUS).… See the full description on the dataset page: https://huggingface.co/datasets/matiasqr/greenery-perception.CWRU-perception
Roles
Roles: perception view of CWRU — annot is the source label (ball / inner_race / normal / outer_race), kept machine-parseable as the gold for verification and reward parsing; the model reads query + image, where the repo ships a bearing's vibration in four image encodings as four equal-sized configs — reshaped (consecutive samples arranged as the rows of a grayscale square), scalogram (a continuous-wavelet time-scale view), spectrogram (a short-time Fourier transform) and… See the full description on the dataset page: https://huggingface.co/datasets/AI4Manufacturing/CWRU-perception.Puzzle_Perception
Puzzle Perception — Segmentation + pVQA
A single table over two tasks on synthetic puzzle images:
Segmentation — per-pixel masks over chess, maze and tower-of-hanoi under one
unified 30-class label space.
pVQA — multiple-choice perception probes over chess and N-Queens boards.
Every row carries the same 13 columns; the type column says which task it
belongs to, and columns that do not apply are null.
from datasets import load_dataset
ds =… See the full description on the dataset page: https://huggingface.co/datasets/PuzzleComm/Puzzle_Perception.IMS-perception
Roles
Roles: perception view of IMS — annot is the source label (ball / inner_race / normal / outer_race), kept machine-parseable as the gold for verification and reward parsing; the model reads query + image, where the repo ships a bearing's vibration in four image encodings as four equal-sized configs — reshaped (consecutive samples arranged as the rows of a grayscale square), scalogram (a continuous-wavelet time-scale view), spectrogram (a short-time Fourier transform) and… See the full description on the dataset page: https://huggingface.co/datasets/AI4Manufacturing/IMS-perception.lighting-invariant-bedroom-perception-robustness-benchmark-next-pack-1917c2cb-f1975230
Home Object Detection, Grasping and Sorting Eval — YOLOv8
Evaluation dataset for a robotic arm that detects, grasps, and sorts objects by type in home environments. 30 renders at 640x640 across kitchen, entry, living room and dressing spaces, staged with everyday household objects. Includes RGB plus metric depth, world-space normals (OpenGL, linear), albedo and material index passes, per-frame annotations, and midday lighting. Targets a YOLOv8 model.
This dataset mirrors public… See the full description on the dataset page: https://huggingface.co/datasets/physicl-community/lighting-invariant-bedroom-perception-robustness-benchmark-next-pack-1917c2cb-f1975230.XJTU-perception
Roles
Roles: perception view of XJTU — annot is the source label (inner_race / normal / outer_race), kept machine-parseable as the gold for verification and reward parsing; the model reads query + image, where the repo ships a bearing's vibration in four image encodings as four equal-sized configs — reshaped (consecutive samples arranged as the rows of a grayscale square), scalogram (a continuous-wavelet time-scale view), spectrogram (a short-time Fourier transform) and waveform… See the full description on the dataset page: https://huggingface.co/datasets/AI4Manufacturing/XJTU-perception.LabHorizon-3D-Asset-Perception
LabHorizon 3D Asset Perception
Pushing the Limits of Laboratory 3D Perception and Long-Horizon Planning via Protocol-Aligned Action Prediction
Overview | News | Highlights | Dataset | Evaluation | Leaderboard | Training | Citation
🔎 Overview
This dataset is the Level 1 split of LabHorizon. Each example pairs three rendered views of the same laboratory asset with historical experimental actions and a set of… See the full description on the dataset page: https://huggingface.co/datasets/Backup-SU-CongLab/LabHorizon-3D-Asset-Perception.MENDELEY-rotor-perception
Roles
Roles: perception-only repo (no reasoning track is published for this dataset) — annot is the source label (looseness / misalignment / normal / unbalance), kept machine-parseable as the gold for verification and reward parsing; the model reads query + image, where the repo ships a rotating machine's vibration in four image encodings as four equal-sized configs — reshaped (consecutive samples arranged as the rows of a grayscale square), scalogram (a continuous-wavelet… See the full description on the dataset page: https://huggingface.co/datasets/AI4Manufacturing/MENDELEY-rotor-perception.V-Perception-40KPHM2010-perception
Roles
Roles: perception view of PHM2010 — annot is the source label (serviceable / worn), kept machine-parseable as the gold for verification and reward parsing; the model reads query + image, where the repo ships a milling spindle's axial cutting force in four image encodings as four equal-sized configs — reshaped (consecutive samples arranged as the rows of a grayscale square), scalogram (a continuous-wavelet time-scale view), spectrogram (a short-time Fourier transform) and… See the full description on the dataset page: https://huggingface.co/datasets/AI4Manufacturing/PHM2010-perception.imaginative-perception-token-pet-eval-ai2thor_filteredOTTAWA-VARSPEED-perception
Roles
Roles: perception view of OTTAWA-VARSPEED — annot is the source label (ball / combined / healthy / inner_race / outer_race), kept machine-parseable as the gold for verification and reward parsing; the model reads query + image, where the repo ships a variable-speed bearing's vibration in four image encodings as four equal-sized configs — reshaped (consecutive samples arranged as the rows of a grayscale square), scalogram (a continuous-wavelet time-scale view), spectrogram… See the full description on the dataset page: https://huggingface.co/datasets/AI4Manufacturing/OTTAWA-VARSPEED-perception.SEUG-perception
Roles
Roles: perception view of SEUG — annot is the source label (gear_fault / healthy), kept machine-parseable as the gold for verification and reward parsing; the model reads query + image, where the repo ships a gearbox's vibration in four image encodings as four equal-sized configs — reshaped (consecutive samples arranged as the rows of a grayscale square), scalogram (a continuous-wavelet time-scale view), spectrogram (a short-time Fourier transform) and waveform (the… See the full description on the dataset page: https://huggingface.co/datasets/AI4Manufacturing/SEUG-perception.PADERBORN-perception
Roles
Roles: perception view of PADERBORN — annot is the source label (healthy / inner_race / outer_race), kept machine-parseable as the gold for verification and reward parsing; the model reads query + image, where the repo ships a bearing's vibration in four image encodings as four equal-sized configs — reshaped (consecutive samples arranged as the rows of a grayscale square), scalogram (a continuous-wavelet time-scale view), spectrogram (a short-time Fourier transform) and… See the full description on the dataset page: https://huggingface.co/datasets/AI4Manufacturing/PADERBORN-perception.