#Japan #Tokyo #AI Data #Qlean Dataset #Visual Bank

Visual Bank Introduces the Speed Skating Dataset

Visual Bank, a startup based in Minato, Tokyo, under the leadership of CEO Masayuki Nagai, is spearheading the release of the Qlean Dataset to enhance AI development tools. Recently, they have added a new Speed Skating Video and Image Dataset to their lineup, expanding their proprietary collection known as AI Data Recipes. This initiative aims to provide versatile data solutions for both academic research and commercial AI development.

The Qlean Dataset Overview

The newly launched Speed Skating Dataset captures high-quality videos and images of male speed skaters, specifically featuring three Japanese athletes in their twenties. The data is collected in diverse shooting environments—specifically indoors at a skating rink—with various angles and shooting distances to provide a comprehensive view suitable for analysis.

Video Dataset Details

• - Data Size: 6.03 GB
• - Number of Items: 18
• - Data Format: MP4
• - Total Filming Duration: 22 minutes and 36 seconds
• - Sample Video Details

Image Dataset Details

• - Data Size: 116.06 GB
• - Number of Items: 19,038
• - Data Format: JPEG
• - Sample Image Details

Potential Use Cases for the Speed Skating Dataset

The Speed Skating Dataset offers numerous applications across various fields:

Sports Motion Analysis and Form Optimization

By providing high-resolution images and multiple angles of skaters in motion, this dataset is ideal for developing form analysis AI and posture estimation algorithms. Coaches can leverage the insights gained to enhance performance metrics and refine training methods.

Behavior Recognition and Motion Tracking AI Research

The dataset includes high frame rate footage of skaters reaching speeds of over 50 km/h, making it valuable for testing behavior recognition AI and motion analysis models. Its unique characteristics allow for research in both sports science and industrial camera technology.

Development of Scoring and Form Evaluation Support Systems

The detailed data allows stakeholders to accurately assess players' body axes and center of gravity, making it useful for scoring assistance and developing form evaluation AI. It particularly aids in automating quantitative assessments in training facilities and federations.

VR/AR Experience Content Development

The multi-angle images and videos lend themselves well to generating realistic skating experiences in VR/AR environments. This capability can be harnessed for educational purposes or promotional content for sports competitions.

Robotics and Motion Mimicry Learning Applications

Capturing crucial data on balance control and center of gravity motion, the dataset can be instrumental in autonomous movement research and robotics control training. It can help build models for stability and speed regulation in robotic systems.

Special Features of the Qlean Dataset

The Qlean Dataset is designed to meet the needs of researchers and commercial enterprises alike. Each dataset is backed by agreements for data utilization from all subjects involved, ensuring compliance with privacy policies across different regions.

Employing the unique structure of AI Data Recipes, Visual Bank maximizes ROI while keeping initial investments down. Moreover, for datasets not included in the AI Data Recipes lineup, custom data sets can be created according to specific requirements, showcasing the company's adaptability.

Academia Support Program

Visual Bank has also initiated a dataset provision program for academia, offering free access to datasets for universities, research institutions, and non-profit technology development teams. They provide over 500,000 data samples across 80 categories, addressing the significant challenges in obtaining high-quality, rights-cleared learning data in research settings. More details are available on their Academia Program page.

About Visual Bank

Visual Bank is dedicated to building and providing next-generation data infrastructure that amplifies AI development capabilities. Their mission is to unlock the potential of all types of data. Alongside their data solutions, they support manga artists with their AI-assisted tool, THE PEN. Having been selected for Japan’s national GENIAC research and development program, Visual Bank is firmly positioned at the forefront of AI innovation and implementation.

CEO: Masayuki Nagai
Location: 107-0062 6F, C-Cube Minami-Aoyama Building, 7-1-7 Minami-Aoyama, Minato, Tokyo
Visual Bank Website
Amana Images Website