Video · Image · 3D / Point cloud · Multimodal / AI & Machine Learning
Egocentric Data | Professionals at Work on Live Sites | Hand Tracking, IMU & Per-Second Action Labels
Advertised by providerLicensing rights not independently verifiedPaid licence
Overview
Egocentric Data | Professionals at Work on Live Sites | Hand Tracking, IMU & Per-Second Action Labels is a multimodal dataset. Typical applications include LLM pre-training and robotics & embodied AI. Relevant sectors: AI & Machine Learning and Manufacturing & Industrial. The listing mentions timestamps. It is offered as a paid dataset; pricing details are provided on request.
Primary applications
What the listing offers
- Listing mentions timestamps.
Azimov is egocentric (first-person) data recorded by trained professionals doing their actual jobs on live, operating sites. It is not staged, scripted or crowdsourced. Workers wear head-mounted capture rigs during normal paid shifts, so the data captures the real tools, real variation, and the errors, retries and recoveries that scripted collections leave out. WHAT IS IN THE DATASET Our first collection, Azimov300, covers facility-maintenance work: 110+ unique tasks across 11+ environments, recorded with 10 technicians on an operating site in Qatar. New task families (logistics and construction) are being added from October 2026, with European manufacturing and printing sites from late November 2026. Each recording includes: - Six synchronised camera feeds (stereo, control and tracking cameras) - Hand tracking, processed in-house - IMU (accelerometer and gyroscope) and head position - An action description every second - A scene description every 70 seconds - Episode-level task descriptions, searchable by task All sensor streams are time-aligned, so teams can segment, search and train without running their own labelling pass first. USE CASES - Pretraining vision-language-action (VLA) and robot foundation models on varied, real manipulation with per-second action language - Dexterous manipulation and hand-object interaction: how skilled workers grasp, reposition and use tools - World models and video prediction: egocentric video tied to body motion through IMU and head pose -
Related datasets
More in Video & Physical AI →Image · 3D / Point cloud · MultimodalAdvertised
Computer Vision Data: Multi-View Drone Imagery & 3D Scenes | Multi-Country
The provider provides commercially licensable multi-view drone imagery and associated reconstructed 3D and geospatial data.
- Language
- —
- Volume
- —
- Domain
- Automotive & Mobility +3
- Use case
- Robotics & embodied AI
USD 10,000
More video & physical ai datasets
All 302 video & physical ai datasets →- Egocentric Hand Manipulation Dataset | Cloth-Folding Episodes with Wrist-Camera Video & IMU | LeRobot v3.0, Global Use
- Egocentric head-mounted video of everyday household tasks
- Egocentric Hotel & Lodging Housekeeping Video Data: Cleaning, Turnover & Laundry | SLAM/VIO Head-Motion Tracking | Embodied AI, Physical AI
- Egocentric Household Activity Demonstrations — LeRobot
- Egocentric human, synthetic, and teleoperated robot video
- Egocentric Multi-View Dataset with 21-points 3D hand Keypoints Annotations
- Egocentric Personal-Care & Salon Video: Haircutting, Manicure & Pedicure | Machine Learning (ML) Data, Synced IMU | Fine-Motor Robotics
- Egocentric POV video of commercial kitchen service workflows