{"success":true,"data":{"author":"OpenGraph Labs","bucketAccessType":"public","bucketName":"OpenGraphLabs-Research/ego-tactile-manipulation","bucketPrefix":"","category":"Robotics","cloudProvider":"huggingface","createdAt":"2026-11-21T22:15:33.144Z","description":"A sample of synchronized egocentric video and hand touch data for learning human manipulation, from OpenGraph Labs' physical-AI data pipeline. A person wore a head camera (1920x1080 at 30 fps, intrinsics included) and OGLO tactile gloves on both hands, each with 5 fingers x 4x4 taxels (80 pressure channels per hand, 160 total, raw at 90 Hz and aligned at 30 Hz), plus a 9-channel head IMU and 6-channel IMUs on both wrists. The set has 72 episodes, 1.28 hours and 138,629 frames of tabletop tasks such as pouring, capping, cutting, gripping and arranging. Action boundaries come from the physical contact and grip-force signal rather than human judgment, and concise English labels for action, sub-action, anchor, grasp, hand and object are attached within each span. Packaged as LeRobot v3.0 (meta/, per-frame Parquet in data/, video in videos/), with byte-exact raw tactile and IMU in extra/ and one Rerun .rrd recording per episode for visualization; about 6 GB (measured 2026-11-21). Load with lerobot's LeRobotDataset or pyarrow. Released under CC-BY-4.0, including commercial use with attribution.","displayTitle":"Ego-Tactile Manipulation: Egocentric Video with 160-Channel Two-Hand Tactile Glove Pressure and Touch-Grounded Action Labels (OpenGraph Labs)","huggingFaceUrl":"https://huggingface.co/datasets/OpenGraphLabs-Research/ego-tactile-manipulation","id":"dbe52fcf-cff8-5232-995b-ec91ae5b4a89","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/OpenGraphLabs-Research/ego-tactile-manipulation","objectCountEstimate":612,"probeStatus":"ok","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"OpenGraphLabs-Research/ego-tactile-manipulation\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"OpenGraphLabs-Research/ego-tactile-manipulation\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=OpenGraphLabs-Research, license:cc-by-4.0, size_categories:1M<n<10M, format:parquet, downloads=1436"],"format":"mixed","generatedAt":"2026-11-21T22:31:32.059Z","layout":{"example":"data/chunk-000/file-000.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"data/chunk-000/file-000.parquet","truncatedAt":512}],"sampleObjects":["assets/action_seg_browser.png","assets/action_seg_storyboard.png","assets/ego_tactile.png","assets/ego_tactile_grid.mp4","assets/oglo_gloves.png"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\nlanguage:\n- en\npretty_name: Ego-Tactile Manipulation\nsize_categories:\n- 100K<n<1M\ntask_categories:\n- robotics\ntags:\n- LeRobot\n- robotics\n- physical-ai\n- egocentric\n- tactile\n- manipulation\n- action-segmentation\n- multimodal\n---\n\n# Ego-Tactile Manipulation\n\n*Egocentric video + dense two-hand tactile + touch-grounded action labels - by [**OpenGraph Labs**](https://www.opengraphlabs.com/).*\n\n<video controls autoplay loop muted playsinline width=\"100%\" src=\"https://huggingface.co/datasets/OpenGraphLabs-Research/ego-tactile-manipulation/resolve/main/assets/ego_tactile_grid.mp4\"></video>\n\n<sub>Four episodes visualized in our dashboard - egocentric video with live tactile & sensor signals.</sub>\n\nSynchronized **ego + touch** human-manipulation data is rare. This is a clean **1.28-hour sample** from OpenGraph Labs' Physical-AI data pipeline: a head camera plus our **OGLO tactile gloves** on both hands, with action labels derived from the physical contact signal.\n\n## OGLO - our tactile gloves\n\n![OGLO tactile gloves](https://huggingface.co/datasets/OpenGraphLabs-Research/ego-tactile-manipulation/resolve/main/assets/oglo_gloves.png)\n\nWorn on both hands · **5 fingers × 4×4 taxels = 80 channels per hand (160 total)** · 90 Hz pressure, co-registered to every video frame.\n\n## Touch-grounded action segmentation\n\nAction boundaries come from physical **contact and grip-force** - no human ground truth. Concise labels are added within each physically-fixed span.\n\n![Action storyboard](https://huggingface.co/datasets/OpenGraphLabs-Research/ego-tactile-manipulation/resolve/main/assets/action_seg_storyboard.png)\n![Action browser](https://huggingface.co/datasets/OpenGraphLabs-Research/ego-tactile-manipulation/resolve/main/assets/action_seg_browser.png)\n\n## Specs\n\n| | |\n|---|---|\n| Episodes / duration | 72 · **1.28 h** (138,629 frames @ 30 fps) |\n| Camera | egocentric, 1920×1080 (intrinsics included) |\n| Tactile | 160 channels · 30 Hz aligned + 90 Hz raw |\n| IMU | head (9-ch) + both wrists (6-ch each) |\n| Annotations | action / sub-action / anchor / grasp / hand / object (English) |\n| Tasks | tabletop manipulation - pour, cap, cut, grip, arrange, … |\n| License | CC-BY-4.0 |\n\n## What's inside\n\n```\nmeta/      LeRobot v3.0 metadata  (+ opengraph/: camera intrinsics, tactile highlights)\ndata/      per-frame parquet      (tactile @30 Hz, IMU, annotations)\nvideos/    egocentric video\nextra/     raw_tactile/ + raw_imu/ (head, both wrists)   byte-exact raw sensors\nrerun/     one .rrd per episode   (instant visualization)\n```\n\n## Format & use\n\nPackaged in **[LeRobot v3.0](https://huggingface.co/docs/lerobot)** and **[Rerun](https://rerun.io)**, so it's ready to use as-is:\n\n```python\nfrom lerobot.datasets import LeRobotDataset\nds = LeRobotDataset(\"OpenGraphLabs-Research/ego-tactile-manipulation\")\n```\n```bash\nrerun rerun/episode_000000.rrd      # visualize an episode\n```\n\n## About OpenGraph Labs\n\nPhysical-AI data infrastructure - we capture and post-process large-scale egocentric human manipulation data (multi-camera, tactile, IMU, metric egomotion, frame-level semantics). This is our first public sample; reach out for collaborations and larger releases.\n\n🌐 [opengraphlabs.com](https://www.opengraphlabs.com/) · 🤗 [OpenGraphLabs-Research](https://huggingface.co/OpenGraphLabs-Research)\n\n## Citation\n\n```bibtex\n@misc{opengraphlabs2026egotactile,\n  title  = {Ego-Tactile Manipulation},\n  author = {OpenGraph Labs},\n  year   = {2026},\n  url    = {https://huggingface.co/datasets/OpenGraphLabs-Research/ego-tactile-manipulation}\n}\n```\n\nReleased under **CC-BY-4.0** - free to use, including commercially, with attribution.\n","encoding":"utf8","key":"README.md","name":"README.md","size":3678,"truncated":false}],"sourceObjectKey":"data/chunk-000/file-000.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"OpenGraphLabs-Research/ego-tactile-manipulation\")\nprint(ds)","starterSamplePath":"assets/action_seg_browser.png","totalObjects":612},"summary":"A sample of synchronized egocentric video and hand touch data for learning human manipulation, from OpenGraph Labs' physical-AI data pipeline.","title":"Ego-Tactile Manipulation: Egocentric Video with 160-Channel Two-Hand Tactile Glove Pressure and Touch-Grounded Action Labels (OpenGraph Labs)","updatedAt":"2026-11-22T00:47:14.011Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/OpenGraphLabs-Research/ego-tactile-manipulation"}}