“We were very happy with the quality of the data that we purchased. It will assist our efforts to improve our product. The team at Unidata was very responsive to our needs, and provided the data format needed to complete our tasks. We are happy to state that Unidata delivers a high quality product in an effective and efficient manner.”
Commercial
Egocentric Video Dataset
This egocentric dataset contains first-person videos of daily activities in home environments, designed for Physical AI, robotic systems, manipulation tasks, and egocentric vision research. The videos were captured using 2 hardware configurations: Pico + Motion Trackers (2,321 hours) and Zed + Pico + Motion Trackers (1,729 hours). Quaternion-based orientation supports 3D pose estimations and egocentric tracking from first-person.
-
- Hours
- 4,050
-
- Scenarios
- 13
- Computer Vision
- Human Activity Recognition
- Robot Learning
- Motion Analysis
- Egocentric Vision
- Computer Vision
- Human Activity Recognition
- Robot Learning
- Motion Analysis
- Egocentric Vision
This egocentric dataset contains first-person videos of daily activities in home environments, designed for Physical AI, robotic systems, manipulation tasks, and egocentric vision research. The videos were captured using 2 hardware configurations: Pico + Motion Trackers (2,321 hours) and Zed + Pico + Motion Trackers (1,729 hours). Quaternion-based orientation supports 3D pose estimations and egocentric tracking from first-person.
- Computer Vision
- Human Activity Recognition
- Robot Learning
- Motion Analysis
- Egocentric Vision
-
- Hours
- 4,050
-
- Scenarios
- 13
Dataset Info
| Characteristic | Data |
| Description | Egocentric video recordings of daily activities in home environments |
| Data types | Video |
| Tasks | Hand Activity Recognition, Egocentric Action Recognition, Hand-Object Interaction |
| Hours of recordings | 4,050 |
| Hardware setups | 2 |
| Setup 1 (Pico + Motion Trackers) | 2,321 hours (57.3%) — natural speed, slow-motion, and real-speed object transferring |
| Setup 2 (Zed + Pico + Motion Trackers) | 1,729 hours (42.7%) — scripted object transfer tasks with spatial depth + egocentric view |
| Scenarios | 13 (sorting unsorted items, arranging products by category, collecting items into a container, transferring from drawer to table, wardrobe & table & bag, transport box & display table, folding fabric items, lids & cookware & drawers, transferring with a spoon, transferring with tongs, packing into containers, two-handed sorting, assembly & disassembly) |
| Environments | Kitchen, bathroom, living room, and other home settings |
| Activities | Sorting, transferring, folding, assembly/disassembly, tool use, two-handed manipulation |
Technical
Characteristics
| Characteristic | Data |
| Video source | Pico 4 Ultra VR headset (egocentric) + 4 Zed stereo cameras (spatial depth) |
| Recording speed | Real-time / Slow-motion |
| Hand visibility | As needed / Both hands always in frame (slow-motion recordings) |
| Extension of labeling file | .txt per recording |
| Orientation format | Quaternions (from onboard sensor-fusion) |
| Sensor data | IMU signals — accelerometer, gyroscope, magnetometer |
Statistics
- Hardware configuration distribution
Before You Buy The Full Set
This dataset is available for AI stress testing. We run your model on it and report error rates per class, segment & condition, so you buy only the 5–30% that fixes the classes it missed.
Show me what fails
Scenarios
| i | Variable | Hours | % |
| 1 | Sorting unsorted items | 800 | 19.8% |
| 2 | Arranging products by category | 800 | 19.8% |
| 3 | Collecting items into a container | 400 | 9.9% |
| 4 | Transferring from drawer to table | 400 | 9.9% |
| 5 | Wardrobe, table & bag | 400 | 9.9% |
| 6 | Transport box & display table | 400 | 9.9% |
| 7 | Folding fabric items | 400 | 9.9% |
| 8 | Lids, cookware & drawers | 200 | 4.9% |
| 9 | Transferring with a spoon | 50 | 1.2% |
| 10 | Transferring with tongs | 50 | 1.2% |
| 11 | Packing into containers | 50 | 1.2% |
| 12 | Two-handed sorting | 50 | 1.2% |
| 13 | Assembly & disassembly | 50 | 1.2% |
| 14 | Total | 4,050 | 100% |
Dataset Use Cases
FAQ
How was the data collected?
Data was captured using two hardware configurations: Pico VR headset with motion trackers, and Zed + Pico + motion trackers, while participants performed daily household activities. Recording conditions vary across subsets, covering real-time speed, slow-motion kinematics, and scripted object transfer scenarios.
What makes this egocentric dataset suitable for robot learning?
Unlike generic action recognition datasets, this collection focuses on daily household manipulation tasks from a first-person perspective using wearable sensors. The combination of egocentric video, IMU signals, and realistic object interactions provides high-quality training data for Physical AI, embodied AI, and robot learning.
Is this dataset unique?
Yes. The combination of slow-motion kinematic recording, dual-camera setup (Zed + Pico), and scripted object transfer scenarios across six structured subsets is not replicated in existing datasets. The multimodal data, such as egocentric video, IMU signals, and quaternion orientation, make it distinct for robotic and Physical AI research.
Is this real-world or synthetic data?
This is entirely real-world data. Participants performed genuine daily household actions, including object transfers across surfaces, giving the dataset the natural human motion variability that synthetic datasets typically lack.
How are Unidata datasets licensed?
Unidata datasets follow a dual-licensing model: free samples are provided for trial and testing, while full datasets are available exclusively through purchase. Licensing terms are confirmed during the request and documentation process.
Do Unidata datasets comply with GDPR?
Yes. All datasets are curated in compliance with GDPR and applicable data protection laws. Data is collected from legally permissible sources to ensure ethical and lawful usage across research and commercial applications.
How are Unidata datasets stored?
All datasets are stored securely on AWS cloud infrastructure, aligned with ISO 27001 and ISO 27701 standards for information security and privacy management. This ensures high availability, scalability, and a privacy-focused environment for handling sensitive visual data.
How long does delivery take?
Once you submit a request, Unidata will contact you to review the details and complete the necessary documents. After signing and payment, the dataset is delivered within 3–10 days.
Still have questions about using Unidata datasets?
Read our user-guides
Unidata Cases
Digital Tree Passport Annotation for Forest Mapping
- Forestry Monitoring & GIS
- 200,000 trees, 10 species classes
- 2 months
License Plate Annotation for Vehicle Recognition System
- 100,000 images with detailed license plate markup (bounding boxes, digits, regional symbols)
- 3 weeks
Sentiment Annotation for Brand Monitoring
- Marketing & Consumer Insights
- 12,000 text samples
- 7 weeks
Surveillance Video Annotation for Entrance Monitoring
- Surveillance & Security
- 90 minutes of video from three cameras, approximately 50-60 thousand frames
- 2 week
Similar Datasets
-
Commercial
- Computer Vision
- 3D Reconstruction
- Mesh Generation
- Gaussian Splatting
- Scene Understanding
3D Residential Scans Dataset
This dataset contains 3D scans of apartments, including splats and meshes generated via 3DGS technology, with high-quality point clouds, lidar scans, RGB-D data, and textured 3D meshes captured using the XGRIDS PortalCam. Designed for robotics, physical AI, and robot training, this real-world 3DGS dataset supports 3D reconstruction, scene understanding, object detection, spatial analysis, and deep learning workflows for indoor environments.
100 Scans
-
Commercial
Lerobot SO-101 Manipulations Dataset
Lerobot SO-101 dataset is a compact, high-quality set of 20 teleoperated robotic arm recordings (~30 FPS) captured on the calibrated robot studio SO-101 arm using multi-view cameras (front, wrist, top) in the Lerobot framework. Designed for imitation learning and model training on the Lerobot platform, it demonstrates pick-and-place and basic manipulation tasks with structured LeRobot v2.1 annotations.
This dataset is intended as a limited reference set and contains 20 episodes; however, full custom data collection in Lerobot format can be provided on request.
20 Episodes
~30 FPS -
Commercial
- Computer Vision
- Machine Learning
- Human Activity Recognition
- Robot Learning
- Motion Analysis
Robotic Household Activities Dataset
This droid dataset includes recordings of cleaning, laundry folding, dishwashing activities, combining head-mounted video with detailed robotic data streams from seven 9-axis IMU units placed on the chest, upper arms, forearms, wrists. The robot dataset captures data from 3-axis accelerometers, 3-axis gyroscope, and 3-axis magnetometer sensors, using onboard sensor-fusion to generate orientation quaternions for motion tracking and robotic learning.
1,000 Hours
Why Companies Trust Unidata's Datasets
Share your project requirements, we handle the rest. Every service is tailored, executed, and compliance-ready, so you can focus on strategy and growth, not operations.
Our Clients Love Us
Ready to get started?
Tell us what you need — we’ll reply within 24h with a free estimate
- Andrew
- Head of Client Success
— I'll guide you through every step, from your first
message to full project delivery
Thank you for your
message
We use cookies to enhance your experience, personalize content, ads, and analyze traffic. By clicking 'Accept All', you agree to our Cookie Policy.