Asilla, a developer of security systems with world-class behavior recognition AI at its core, launched a dedicated team for human science research, the Human Science AI Research Team, in February 2023.
This article interviews Mr. Masahiro Wakasa, CTO of Asilla, the founder of the team, to find out how "gait analysis," one of Asilla's core technologies, and HSAR research are connected.
Gait analysis is a technology that focuses on the so-called "human gait. It is a field that requires in-depth knowledge of the human gait, as it is the research and development of technology that identifies "who the person is" based on characteristic movements.
Currently, Asilla is conducting two main types of research in gait analysis: "gait recognition" and "trajectory analysis.
The "gait recognition" is a technology that uses the individuality of a person's walking style to identify them. It is the only technology that can authenticate an individual by recognizing the way they walk, even when their face cannot be identified in the video.
For example, with a security camera, it may be difficult to identify a person because the face cannot be recognized depending on the distance or angle of view, but the Walking Pattern Recognition System can identify a person because it can recognize the way he/she walks even from a distance of 50 to 100 meters. Another feature of the system is that it can identify and track a person even between different cameras because it focuses on the way he or she walks.
As a biometric authentication technology that uses human physical characteristics to authenticate individuals, "gait recognition" is also attracting attention for its potential to become as accurate as facial recognition in the future. It is also expected to be used in crime prevention efforts and layout design in urban development.
Trajectory analysis" conducted by Asilla is a study to identify a person's psychological state and to infer possible future behavior based on the path (trajectory) taken by the person.
For example, normal people tend to walk in a straight line, but in the case of a person whose movements are different from general movements, such as a tissue distributor or a catch-salesman, the person may move back and forth in the same area many times or talk to others, which creates differences in their movements. In Trajectory Analysis, the approach is for AI to automatically detect and extract such differences in movement.
Another difference is that "gait analysis" focuses on "gait" and has a short time axis of only a few seconds, while "trajectory analysis" focuses on "paths" and targets long-term movements of around 10 seconds. In the AI industry, there is an algorithm called "Anomaly detection," which distinguishes between normal and abnormal behavior, and we are currently researching it and considering its application to our products.
Asilla is thus deepening its research on "gait analysis" from the two approaches of "gait recognition" and "trajectory analysis," and is promoting research and development to further improve the accuracy of AI detection and inference.
As reported in the article on the launch of HSAR, in gait analysis, Asilla is conducting research from two axes: traditional "computer science" and "human technology" to deepen our understanding of people themselves.
Until now, our research has focused on the "typical computer science approach," in which AI is given learning data on various human walking styles, and AI learns automatically. However, there was a concern that this approach might eventually reach a plateau, even though it could improve accuracy to some extent, because of the risk of the AI automatically learning with incorrect parameters.
Therefore, HSAR incorporates the knowledge of human science into the process of "assigning the points (parameters) that AI should originally look at" from this perspective. It is like the findings of human science becoming "the right guideposts for AI.
For example, for a male in his early 30s, his walking speed is about 5.7 km/hour, and the average stride length is about 70 cm. By adapting such "parameters to be viewed from human science" to AI, Asilla aims to further improve the accuracy of AI.
Although it is still only a sensory value, I feel that we are already beginning to see results. One of the most significant achievements is that "what used to be a black box inside AI is gradually becoming visible.
What I mean by this is that AI has been complicated in order to improve its forecasting accuracy, and thus has been a problem of "black-boxing," in which the criteria and basis for decisions are unclear. However, with the addition of a human science perspective to AI, it is now possible to understand the axis and basis of AI decisions, such as "this AI paid attention to this part of the way you walk.
For example, AI sometimes made judgments by looking at the head, even though it should have done so by looking at the angle of the elbow. After that, we pointed out the wrong point to the AI and gave it a "viewpoint (parameter) to be looked at from a human science perspective," which improved the results and accuracy.
Such "visualization of AI" has led to the discovery of a promising model that is consistent with the viewpoints (parameters) of both human science and AI, or, conversely, to the discovery of a model to be reevaluated that is judged based on parameters that are completely different from those that AI should be looking at. In other cases, we can find models to be reevaluated that are judged by looking at parameters that are completely different from those that AI should be looking at.
While the introduction of human science is expected to further improve the accuracy of AI, we believe that there is a possibility that AI beyond the understanding of human science will emerge. For example, even though the head is not considered an important parameter in current human science, AI may lead to new discoveries that have not been considered in existing research, such as the fact that the way the head moves has a significant impact on age and gender.
In fact, the phenomenon of "AI surpassing humans" has already occurred in other industries, with a familiar example being recommendation systems. The same may be true for human behavior, which Asilla is working on. It is exactly as if AI is going to surpass humans.
While keeping such a possibility in mind, Asilla does not believe that "human science is 100% correct," but rather aims to "develop AI with a good balance between humans and AI" by utilizing "human science as knowledge to improve the accuracy of AI. We aim to "develop AI that balances both humans and AI.
Based on the philosophy of "Technology Driven Future," Asilla conducts research and development on a daily basis in order to realize a world in which all people can live in safety and security. In the field of behavior recognition AI, in particular, we have a strong desire to become a world leader, and our researchers and members are working together as a team to constantly evolve.
In the future, we will further promote the further evolution of our behavior recognition AI technology and strive to create an environment where all our engineers can pursue their own careers and always be excited about their work. If you are interested in AI related to human behavior, I look forward to working with you to create something together.
Inquiries about HSAR: pr@asilla.jp
Contact: Wakasa
After completing his graduate studies at Tokyo Institute of Technology in 2018, he joined JGC Corporation. Kuwait, where he was stationed at construction sites and engaged in plant design IT work. He then joined Asilla Corporation in 2020, where he was in charge of product development related to action recognition AI, and was appointed to oversee the IVA Solutions Division, becoming Executive Officer CTO in April 2022 and Director CTO in March 2023.
