Human Science AI: Aiming to Develop AI that Takes on World’s Challenges

Asilla, a company renowned for its exceptional AI technology for behavior recognition, has formed a dedicated research team called the Human Science AI Research Team. This team's primary focus will be on human science. We had the opportunity to interview Mr. Masahiro Wakasa, the Chief Technology Officer of Asilla and the founder of this team, to understand why a company with cutting-edge AI technology is now venturing into human science research.

In the initial installment of this series, we explore the connection between human science and AI, and question why human research is significant in an AI company. We interviewed him to gain insights into this association.

Asilla Inc. CTO Mr. Masahiro Wakasa 

What is the reason behind an AI company delving into human research, and what motivated the establishment of the Human Science AI Research Team (HSAR)?

――What is the rationale behind conducting human research at an AI company, and what motivated the creation of the Human Science AI Research Team (HSAR)?"

Even before behavior recognition AI became widespread, we recognized its potential and started developing our own technology. Our AI and deep learning technology has been primarily focused on recognizing human behavior, but we acknowledge that there are limitations to relying solely on AI. In order to unlock the full potential of behavior recognition AI, it's crucial to not only improve the technology itself, but also to integrate human understanding and expertise with AI. 

In light of this, Asilla has established the Human Science AI Research Team (HSAR) to merge the two domains of AI technology and human science and elevate its self-developed behavior recognition AI to new heights. The HSAR was founded with the goal of advancing Asilla’s proprietary behavior recognition AI to a new stage within the broader field of AI.

-- Could you explain how this differs from the previous approach?

AI primarily analyzes and draws conclusions from historical data. Nonetheless, even with accurate data, AI may sometimes generate outcomes that deviate from genuine human behavior. In situations where traditional AI produces outcomes that seem unnatural, incorporating human knowledge can serve as a corrective measure by providing feedback such as 'that's incorrect; it should be this way.

We are convinced that the integration of conventional AI technology with human insight and knowledge will enable us to capture human-like movements that were previously unattainable. We believe that an understanding of human behavior will enhance the precision of behavior recognition AI even further.

Research that emphasizes human beings is being advanced, with a particular focus on topics such as walking and behavior.

-- What areas of research would you focus on?

The Human Science AI Research Team (HSAR) undertakes a range of studies that center on "aspects connected to human behavior."

◆'Gait analysis' is the process of identifying individuals based on their unique walking patterns.

Allow me to present some examples, beginning with the first one - research on 'gait analysis.'

This technological research and development entail identifying an individual based on their distinctive movements, particularly their walking patterns, requiring a profound understanding of the mechanics of human motion.

Unlike conventional AI, which involves (1) possessing data and (2) memorizing that data, Asilla’s approach centers on analyzing the characteristics of human gait and identifying the types of data that can be extracted through gait analysis.

The approach prioritizes learning techniques that teach AI systems how to absorb human data and parameters, facilitating their ability to learn and adapt accordingly.

◆Analyzing a person's psychological state from their 'human trajectory'

The second research area is 'human trajectory Analysis.' This study concerns the analysis of the paths that individuals take, with a focus on their movements and corresponding patterns.

Typically, people walk in a straight line when they know where they are going. Yet, when they are lost or feel threatened, as in an accident or incident, their trajectory may become less predictable and meandering. Their movements can vary depending on the situation and their emotional state.

By employing these methods, it is possible to conduct research that can determine an individual's psychological state by examining their walking path and trajectory.

- The integration of gait analysis and walking trajectory analysis could pave the way for multiple potential solutions in the future, such as the early detection and prevention of criminal activities.

Asilla’s objective is not to create a surveillance tool but rather to enhance the value of spaces. Through the use of AI and camera image, our aim is to provide people with a tool that can improve the safety, security, and comfort of spaces. Our research prioritizes the goal of enhancing spatial value, rather than surveillance, and we are committed to this approach.

Current Page:【Human Science AI: Aiming to Develop AI that Takes on World’s Challenges】

Next Page:【AI Field and its Different Specializations: Aiming to Develop AI that Takes on World’s Challenges 


Information about HSAR can be found at the following link: 

Human Science AI: Aiming to Develop AI that Takes on World’s Challenges

●AI Field and its Different Specializations: Aiming to Develop AI that Takes on World’s Challenges

●Why Human Science AI Research Team is Required - What is needed for the future of AI?

●Demonstrating towards World's No.1 Behavior Recognition AI Technology

●The Era of AI trained by Generated Data - The Potential for Research Hidden in Generative AI


 [問い合わせ先] 本件に関するお問い合わせはこちらまでどうぞ  

 [募集] 採用情報はこちら