{"success":true,"data":{"author":"Dexora team","bucketAccessType":"public","bucketName":"Dexora/Dexora_Real-World_Dataset","bucketPrefix":"","category":"Robotics","cloudProvider":"huggingface","createdAt":"2026-11-14T14:15:10.685Z","description":"The Dexora real-world dataset is a corpus of teleoperated bimanual dexterous manipulation released with Dexora, an open vision-language-action model for high-DoF bimanual dexterity. The card reports 12.2K episodes, 2.92 million frames and 40.5 hours, collected with a hybrid teleoperation system that uses an exoskeleton for the arms and an Apple Vision Pro for the hands to drive a 36-DoF bimanual robot. Each episode has synchronised RGB from an egocentric head camera, left and right wrist cameras and a static third-person camera, plus robot proprioception, actions and language instructions; the card says episodes follow the LIBERO-2.1 standard. Tasks span pick-and-place (about 55 percent), dexterous manipulation (20 percent, such as twisting caps and cutting leeks), assembly (15 percent) and articulated objects (10 percent), with 347 objects in 17 categories and a published object inventory. The repo has a task-level view under dexora/ (about 200 task folders) and category folders airbot_pick_and_place, airbot_assemble, airbot_articulation and airbot_dexterous. Measured 2026-11-14 at about 214 GB. The card carries an MIT licence badge; cite the Dexora project.","displayTitle":"Dexora Real-World Dataset: High-DoF Bimanual Dexterous Teleoperation","id":"3c576e3e-8437-5ec2-bf7e-61f5abd0e445","license":"MIT","licenseEvidenceUrl":"https://huggingface.co/datasets/Dexora/Dexora_Real-World_Dataset","objectCountEstimate":100000,"probeStatus":"ok","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"Dexora/Dexora_Real-World_Dataset\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"Dexora/Dexora_Real-World_Dataset\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=Dexora, license:mit, downloads=49767"],"format":"mixed","generatedAt":"2026-11-14T14:15:21.749Z","layout":{"example":"airbot_articulation/data/chunk-000/episode_000000.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"airbot_articulation/data/chunk-000/episode_000000.parquet","truncatedAt":512}],"sampleObjects":["airbot_articulation/data/chunk-000/episode_000000.parquet","airbot_articulation/data/chunk-000/episode_000001.parquet","airbot_articulation/data/chunk-000/episode_000002.parquet","airbot_articulation/data/chunk-000/episode_000003.parquet","airbot_articulation/data/chunk-000/episode_000004.parquet"],"sidecars":[{"bytes":"---\nlicense: mit\n---\n<p align=\"center\">\n  <h2>Dexora: Open-Source VLA for High-DoF Bimanual Dexterity</h2>\n</p>\n\n<p align=\"center\">\n  <a href=\"#\"><img src=\"https://img.shields.io/badge/arXiv-2026.xxxxx-B31B1B.svg\" alt=\"arXiv\"></a>\n  <a href=\"https://github.com/ZZongzheng0918/Dexora?tab=readme-ov-file\"><img src=\"https://img.shields.io/badge/Project-Page-blue.svg\" alt=\"Project Page\"></a>\n  <a href=\"LICENSE\"><img src=\"https://img.shields.io/badge/License-MIT-green.svg\" alt=\"License\"></a>\n</p>\n\n\n## 🔥 News & Updates\n\n\n- **2025-12-03**: Released the full **Real-World Dataset** (**12.2K episodes**) on [Hugging Face](https://huggingface.co/datasets/Dexora/Dexora_Real-World_Dataset).\n- **2025-12-12**: Released the **task-level** view of the **Real-World Dataset** (one folder per high-level task) on [Hugging Face](https://huggingface.co/datasets/Dexora/Dexora_Real-World_Dataset).\n- **Coming soon**: We will open-source the full **100K-episode simulation dataset**.\n\n--\n\n## 📊 Dataset Overview\n\nThe Dexora corpus combines **high-fidelity real-world teleoperation data** with a **large-scale simulated dataset** designed to match the embodiment of the physical robot.\n\n### A. Dexora Real-World Dataset (High-Fidelity)\n\nThe Dexora real-world dataset consists of **12.2K teleoperated episodes**, **2.92M frames**, and **40.5 hours** of data. Demonstrations are collected using a **hybrid teleoperation system** that couples an **Exoskeleton** (for arm control) with **Vision Pro** (for dexterous hand control), enabling precise 36-DoF bimanual manipulation on real hardware.\n\n<p align=\"center\">\n  <img src=\"assets/image/dataset.gif\" alt=\"Dexora Multi-view Dataset\" width=\"100%\">\n</p>\n\n<p>\n  <i>Video 1. <b>Synchronized Multi-View Recordings.</b> High-resolution streams from four synchronized views — an ego-centric head-mounted camera, left and right wrist-mounted cameras, and a static third-person scene camera — synchronized with 36-DoF robot proprioception.</i>\n</p>\n\n<p align=\"center\">\n  <img src=\"assets/image/real-data.JPG\" alt=\"Dexora Real-World Dataset Mosaic\" width=\"100%\">\n</p>\n\n<p>\n  <i>Fig 1. <b>High-Fidelity Real-World Scenes.</b> Collected via our hybrid teleoperation system (Exoskeleton for arm + Vision Pro for hand), this dataset covers <b>347 objects</b> across diverse environments. It captures varying lighting conditions, background clutter, and precise bimanual interactions essential for robust policy learning. Panels (a–d) correspond to four task categories: <b>pick-and-place</b>, <b>assembly</b>, <b>articulation</b>, and <b>dexterous manipulation</b>.</i>\n</p>\n\n<p align=\"center\">\n  <img src=\"assets/image/Categorized_Robot_Task_Trajectory_Distribution.png\" alt=\"Dexora Task Categories\" width=\"120%\">\n</p>\n\n<p align=\"center\">\n  <img src=\"assets/image/Robot_Arm_Task_Trajectory_Distribution.png\" alt=\"Dexora Robot Arm Trajectory Distribution\" width=\"120%\">\n</p>\n\n<p>\n  <i>Fig 2. <b>Task Categories & Action Distribution.</b> Unlike standard gripper datasets, Dexora emphasizes high-DoF dexterity. The real-world data distribution includes <b>Dexterous Manipulation (20%)</b> (e.g., <i>Twist Cap</i>, <i>Use Pen</i>, <i>Cut Leek</i>) and <b>Assembly (15%)</b> (e.g., <i>Separate Nested Bowls</i>, <i>Stack Ring Blocks</i>), in addition to <b>Articulated Objects (10%)</b> and <b>Pick-and-Place (55%)</b>.</i>\n</p>\n\nBoth the episodes and annotations follow the **LIBERO-2.1 standard**, including synchronized **RGB observations**, **robot proprioception**, **actions**, and **language instructions**.\n\n### Object Inventory & Reproducibility\n\nReproducibility is a **core value** of Dexora. To enable other labs and industry teams to **faithfully recreate** our environments, we release a **curated object inventory** that mirrors the physical setup used in our real-world experiments.\n\n- **Scale**: **347 objects** across **17 semantic categories** (e.g., tools, containers, articulated objects, deformables, daily-use items).\n- **Coverage**: Objects are chosen to stress **dexterous control**, **bimanual coordination**, and **long-horizon manipulation**.\n-","encoding":"utf8","key":"README.md","name":"README.md","size":17849,"truncated":true}],"sourceObjectKey":"airbot_articulation/data/chunk-000/episode_000000.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"Dexora/Dexora_Real-World_Dataset\")\nprint(ds)","starterSamplePath":"airbot_articulation/data/chunk-000/episode_000000.parquet","totalObjects":100000},"summary":"The Dexora real-world dataset is a corpus of teleoperated bimanual dexterous manipulation released with Dexora, an open vision-language-action model for high-DoF bimanual…","title":"Dexora Real-World Dataset: High-DoF Bimanual Dexterous Teleoperation","updatedAt":"2026-11-14T17:47:02.758Z","visibility":"public","websiteUrl":"https://github.com/ZZongzheng0918/Dexora"}}