Visual Bank Launches Qlean Dataset for AI Research and Development

Visual Bank Introduces Revolutionary Qlean Dataset

Visual Bank Co., Ltd., located in Shibuya, Tokyo, has launched an exciting addition to its AI data solutions via its subsidiary, Amana Images. The newly introduced Qlean Dataset aims to support diverse research and commercial AI development projects. The current release focuses on an extensive video dataset featuring everyday actions such as walking, running, sitting, and standing, enriched with smartphone usage. This addition is part of Visual Bank’s broader initiative to expand their unique inventory of AI training data known as the AI Data Recipe.

Overview of Qlean Dataset

The Qlean Dataset provides a robust resource featuring video content that is readily useful for AI model training. This particular dataset consists of:

- Data Type: Video
- Attributes: Two male subjects in their 20s
- Total Size: 106 GB
- Count: 720 videos
- Format: mp4

Filming Details

The videos were captured in an indoor gym setting from twelve different angles to provide a multi-perspective view of various actions. Overall, the dataset encapsulates 30 different patterns of everyday movements, ranging from the standard activities of walking or running to varied smartphone operations.

For more detailed information, you can visit here.

Applications of the New Dataset

The Qlean Dataset is designed to support a multitude of applications in the realm of AI and technology. Here are some prominent use cases:

1. Action Recognition AI Base Training: With its collection of everyday motions combined with smartphone usage, this dataset serves as foundational data for action recognition and motion classification AI models.

2. Improving Pose Estimation Models: By capturing transitions between sitting, standing, walking, and stopping from various viewpoints, this dataset proves beneficial for validating accuracy in skeletal and pose estimation algorithms.

3. Smartphone Detection and Accident Prevention AI: Utilizing the data for understanding how individuals walk while using smartphones can lead to the development of systems aimed at detecting and mitigating smartphone-related incidents in busy environments like trains or shopping centers.

4. Behavioral Analysis in Crowded Environments: The multi-angle footage allows the training of AI models to analyze human movement patterns within densely populated settings.

5. Contributions to Autonomous Driving and Surveillance AI: Detecting movements such as stopping or starting while using a smartphone can aid the development of pedestrian prediction systems for autonomous vehicles and monitoring systems for identifying risky behaviors.

6. Healthcare Monitoring and Behavior Assessment Research: Data from everyday actions can be utilized for evaluating activity levels and monitoring lifestyle habits, especially aiding in the care of elderly populations or patients through AI systems.

7. VR/AR Simulations: The multi-view recordings can facilitate simulating actions in VR environments or real-time behavior display in AR applications, enhancing training programs or human interaction studies.

8. Camera Technology Development: The dataset provides valuable data to test and improve autofocus and tracking capabilities in media technologies by covering dynamic movements and positional changes from various angles.

Features of Qlean Dataset

Qlean Dataset stands out as it supports commercial usage with clarity on rights management, making it an ideal choice for research and business applications.

- All necessary agreements have been obtained from the subjects featured, ensuring compliance with various privacy policies across countries, thus granting users peace of mind when utilizing the data.
- The AI Data Recipe model allows for speedy access to datasets, maximizing return on investment by facilitating budget-conscious data procurement.
- Unique data needs can also be met through tailored dataset construction, utilizing Qlean Dataset’s capabilities for custom requirements.

Academia Support Program

To bolster academic research, Visual Bank has initiated a free data provision program targeting universities, research institutions, and non-profit technical development teams. Over 50,000 data points across more than 80 types, including images, audio, videos, and texts, will be provided. As a GENIAC-sponsored company, Visual Bank is committed to solving issues related to sourcing high-quality and rights-cleared learning data for research purposes. More details can be found here.

In a rapidly advancing technological environment, Visual Bank is dedicated to maximizing AI development capabilities by crafting next-generation data infrastructure with the mission of uncovering the potential of all data.

For further inquiries regarding the Qlean Dataset and its applications, please refer to the following websites: