Datasets:
pretty_name: Generic Egocentric Video Dataset
tags:
- egocentric-video
- first-person-video
- video-dataset
- activity-recognition
- action-recognition
- video-annotation
- computer-vision
- machine-learning
- artificial-intelligence
- robotics
- industrial-ai
- manufacturing
- video-analytics
- commercial-environments
- residential-environments
- mp4
license: cc-by-4.0
Generic Egocentric Video Dataset
Overview
The Generic Egocentric Video Dataset is a collection of first-person videos with activity/action labels across commercial and residential environments. The dataset overview reports a total of 228,189 hours of video, 132 tasks, and coverage across 11 countries.
The dataset includes two sample collections:
- Commercial Egocentric Dataset Samples
- Residential Egocentric Dataset Samples
The accompanying geographic and industry distribution overview lists activities across multiple countries and industry sectors, including manufacturing, automotive operations, food services, construction, retail, packaging, and warehousing.
This dataset may be relevant to organizations and research teams exploring first-person video understanding, activity recognition, computer vision, robotics, and video analytics.
Dataset Highlights
- 228,189 hours of reported total video duration.
- Activity/action labels for both commercial and residential samples.
- MP4 video format.
Industry-wise Dataset Distribution
The supplied overview presents the following industry sectors and associated video-duration distribution.
| Industry Sector | Total Duration (Hours) | Distribution (%) |
|---|---|---|
| Agriculture, Landscaping & Miscellaneous | 770 | 0.34% |
| Arts, Crafts & Handmade Production | 2,500 | 1.10% |
| Automotive, Repair & Retail Operations | 17,004 | 7.45% |
| Construction, Facilities Management & Trades | 9,000 | 3.94% |
| Consumer Electronics & Wearables Manufacturing | 1,801 | 0.79% |
| Domestic Services & Home Support | 5,456 | 2.39% |
| Electronic Components, Batteries & PCB Assembly | 12,908 | 5.49% |
| Facility Cleaning & Maintenance | 3,260 | 1.43% |
| Food Preparation & Kitchen Support | 13,753 | 6.03% |
| Food Service, Culinary & Domestic Services | 19,865 | 8.71% |
| General Manufacturing Support & Material Handling | 2,500 | 1.10% |
| General Manufacturing, Metalworking & Plastics | 62,659 | 27.46% |
| Manufacturing Facilities & Factory Operations | 45,655 | 20.01% |
| Optical Products Manufacturing & Inspection | 16 | 0.01% |
| Packaging, Logistics & Warehousing | 6,491 | 2.84% |
| Quality Inspection & Finishing | 62 | 0.03% |
| Retail & Customer Service | 4,475 | 1.96% |
| Textile, Apparel & Accessories Manufacturing | 15,051 | 6.60% |
| Toy Manufacturing | 362 | 0.16% |
| Warehousing & Inventory Management | 5,000 | 2.19% |
Dataset Features
| Feature | Confirmed Information |
|---|---|
| Data modality | First-person (egocentric) video |
| Video format | MP4 |
| Annotation type | Activity/action labels |
| Sample environments | Commercial and residential |
| Total duration | 228,189 hours |
Dataset Structure
The supplied folder screenshot confirms the names of the two sample collections. It does not establish the complete directory layout or the location of individual video and annotation files.
The following is a conceptual representation only:
Generic_Egocentric_Video_Dataset/
├── Commercial Egocentric Dataset Samples/
│ └── [MP4 sample files]
└── Residential Egocentric Dataset Samples/
└── [MP4 sample files]
AI Tasks Supported
The dataset's first-person video content and activity/action labels may be relevant to research and development involving:
- Egocentric video understanding.
- Human activity and action recognition.
- First-person video analysis.
- Computer vision and video understanding.
- Video-based machine learning.
- Robotics and embodied AI research.
- Commercial and residential activity analysis.
- Industrial video analytics research.
Applications
Potential applications include:
- AI and machine learning research: Investigating models that learn from first-person video and activity labels.
- Computer vision: Exploring video-based activity recognition and visual understanding.
- Robotics: Researching visual perception in task-oriented environments.
- Video analytics: Exploring automated analysis of recorded activities.
- Commercial environment analysis: Researching activity patterns in commercial settings.
- Residential environment analysis: Investigating activity recognition in residential settings.
- Academic research: Studying egocentric video understanding across different environments.
Industries
The supplied industry distribution identifies sectors including:
- General manufacturing, metalworking, and plastics.
- Manufacturing facilities and factory operations.
- Food service, culinary, and domestic services.
- Automotive, repair, and retail operations.
- Textile, apparel, and accessories manufacturing.
- Food preparation and kitchen support.
- Electronic components, batteries, and PCB assembly.
- Construction, facilities management, and trades.
- Packaging, logistics, and warehousing.
- Retail and customer service.
- Agriculture and landscaping.
- Facility cleaning and maintenance.
- Consumer electronics and wearables manufacturing.
- Additional sectors listed in the industry distribution table.
Intended Audience
The dataset may be relevant to:
- AI and machine learning researchers.
- Computer vision and video-understanding teams.
- Robotics and embodied AI developers.
- Manufacturing and industrial AI teams.
- Academic institutions.
- Video analytics developers.
- Enterprise organizations evaluating egocentric video data.
Dataset Statistics
| Metric | Reported Value |
|---|---|
| Total video duration | 228,189 hours |
| Video format | MP4 |
| Sample environments | Commercial and residential |
| Annotation type | Activity/action labels |
Advantages
- Large reported total video duration.
- Geographic coverage spanning 11 countries.
- Commercial and residential sample collections.
- Activity/action labels for both collections.
- Industry-wise distribution information.
- Potential relevance to computer vision, AI/ML, robotics, and video analytics.
- A basis for evaluating first-person video data for research and development.
Limitations and Notes
- Statistics are transcribed from the supplied screenshots and have not been independently validated.
- Video resolution, frame rate, recording equipment, and collection methodology are unspecified.
- The detailed annotation schema and label definitions are not yet documented here.
- The number of individual videos and annotated segments is unknown.
- Public availability or download permission for the samples has not been established.
- Licensing, redistribution, and commercial-use permissions require confirmation.
- The industry distribution does not establish that every industry is represented in both the commercial and residential collections.
Summary
The Generic Egocentric Video Dataset is a reported 228,189-hour collection spanning 11 countries and 132 tasks. It includes commercial and residential egocentric video sample collections with activity/action labels, with MP4 confirmed as the video format.
Data Creation
Procured through formal agreements and generated in the ordinary course of business.
Considerations
This dataset is provided for research and educational purposes only. It contains only sample data. For access to the full dataset and enterprise licensing options, please visit our website InfoBay.AI or contact us directly.
-Ph: +91 9236398619
-Email: datareq@infobay.ai