InfoBayAI/Stereo_IMU_Egocentric_Dataset

Stereo IMU Egocentric Dataset

Overview

The Stereo IMU Egocentric Dataset is a multimodal dataset containing egocentric video recordings and Inertial Measurement Unit (IMU) data across multiple industry-related environments.

The dataset is organized into four categories: Automobile, Construction, Garments, and Workshop.

Each category contains stereo video files and an IMU data file. The video content is provided in MP4 format, while the IMU data is provided in CSV format.

The dataset includes left, right, and combined video files, alongside corresponding IMU data files. It is intended to support research and development involving egocentric video understanding, computer vision, sensor-based analysis, spatial reasoning, and robotics.

A sample video, Both video.mp4, is provided to allow users to explore the dataset's video content.

Dataset Highlights

• Egocentric Video Data: First-person-perspective video recordings.
• Stereo Video Content: Left, right, and combined video files.
• IMU Sensor Data: Sensor data provided in CSV format.
• Four Industry Categories: Automobile, Construction, Garments, and Workshop.
• Multiple Data Modalities: Video and sensor data organized within the same category folders.
• Sample Availability: A combined video sample is provided as Both video.mp4.

Dataset Features

Feature	Description
Dataset name	Stereo IMU Egocentric Dataset
Dataset type	Egocentric video and IMU sensor data
Video perspective	Egocentric
Video configuration	Left, right, and combined video files
Video format	MP4
Sensor data format	CSV
Sensor data file	imu.csv
Industry categories	Automobile, Construction, Garments, Workshop
Sample file	Both video.mp4

Industry and Environment Coverage

The dataset is organized into the following four categories:

1. Automobile

Egocentric video and IMU data organized under the Automobile category.

2. Construction

Egocentric video and IMU data organized under the Construction category.

3. Garments

Egocentric video and IMU data organized under the Garments category.

4. Workshop

Egocentric video and IMU data organized under the Workshop category.

Dataset Structure

The following structure reflects the folders and files shown in the provided dataset screenshots.

Stereo_IMU_Egocentric_Dataset/
│
├── Automobile/
│   ├── Both video.mp4
│   ├── imu.csv
│   ├── Left video.mp4
│   └── Right video.mp4
│
├── Construction/
│   ├── Both video.mp4
│   ├── imu.csv
│   ├── Left video.mp4
│   └── Right video.mp4
│
├── Garments/
│   ├── Both video.mp4
│   ├── imu.csv
│   ├── Left video.mp4
│   └── Right video.mp4
│
└── Workshop/
    ├── Both video.mp4
    ├── imu.csv
    ├── Left video.mp4
    └── Right video.mp4

Each category contains the same four file types:

File	Format	Description
Both video.mp4	MP4	Combined video file
Left video.mp4	MP4	Left video file
Right video.mp4	MP4	Right video file
imu.csv	CSV	IMU sensor data file

Sample Dataset

A sample video is provided for users who want to explore the dataset before accessing the complete collection.

Available sample:

• Both video.mp4

The sample represents the combined video file. The other video and sensor files are part of the dataset structure but are not included in the sample selection specified for this release.

Supported AI and Research Areas

The dataset's modalities and subject areas may be relevant to research involving:

• Egocentric video understanding
• Stereo video analysis
• Computer vision
• Visual perception
• IMU-based sensor data analysis
• Multimodal learning
• Spatial reasoning
• Human activity and action understanding
• Robotics and embodied AI
• Sensor-based perception systems

These are potential research applications; specific ground-truth labels and benchmark tasks have not been confirmed.

Applications

Potential applications include:

• First-person video analysis
• Stereo vision research
• Multimodal video and sensor-data research
• Robotics and embodied AI development
• Spatial perception research
• Industrial activity analysis
• Sensor-assisted visual understanding
• Machine learning experimentation
• Computer vision system development

Industries

The dataset is organized into four industry categories:

• Automobile
• Construction
• Garments
• Workshop

These categories represent the folder organization confirmed for this dataset.

Intended Audience

The dataset may be relevant to:

• AI and machine learning researchers
• Computer vision researchers and developers
• Robotics teams
• Autonomous systems researchers
• Multimodal AI developers
• Sensor-data researchers
• Academic institutions
• Industrial AI and research teams
• Enterprise AI development organizations

File Formats

The dataset uses the following confirmed file formats:

Extension	Data Type
.mp4	Video
.csv	IMU sensor data

Summary

The Stereo IMU Egocentric Dataset combines egocentric stereo video files and IMU sensor data across Automobile, Construction, Garments, and Workshop categories.

With left, right, and combined MP4 video files alongside CSV-based IMU data, the dataset provides multimodal material relevant to computer vision, stereo video analysis, sensor-based perception, robotics, and embodied AI research.

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