Machine learning is one of the most important innovations today. In fields ranging from medicine to urban planning, machine learning allows computers to learn complex and subtle patterns that would be difficult for even the best human experts.
Seeing the great potential of this growing field, the Ateneo Institute for Intelligent Visual Environments (ALIVE) is keen to co-develop machine learning solutions with leading experts in various fields.
One of the most surprising things about machine learning is that, despite how powerful computers are, they don’t learn the way humans do. Young children can easily recognize familiar faces, tell when something looks unusual, and make sense of a crowded playground with little guidance. But for computers, the same task can be difficult and arduous. Computer vision systems typically require large datasets, careful labeling, iterative training, and continuous testing before they can adapt to changes in lighting, camera angles, weather, and real-world noise.
This counterintuitive gap—that machines have better perceptual abilities than humans, but require more extensive training than the latter—was the central theme of the second Ateneo Breakthrough Lecture, held on February 26, 2026 in Escaler Hall, where computer scientist Patricia “Pie” Angela R. Abu, Ph.D., presented “Smarter Sight: Building Intelligent Visual Systems for Public Good.”
Abu’s full lecture can be viewed at: ateneo.edu/breakthrough
In his talk, Abu explained why interdisciplinary partnerships are important. Building reliable machine learning systems requires bridging messy reality and mathematical models and proving that the system works under real-world conditions.
An associate professor and dean of the Ateneo de Manila University School of Information Systems and Computer Science (DISCS), Abu leads the team at ALIVE to develop machine learning approaches in computer vision, image processing, and related methods, ranging from biomedical imaging to transportation systems.
In the healthcare field, ALIVE has worked on tools such as a dental image support system and a patch-based deep learning model for detecting bone metastases. This is an example of how machine learning can help experts work more consistently by highlighting patterns in images that are difficult to spot in large images. Another example of ALIVE’s work is V-PROBE (Real-Time Vehicle and Pedestrian Observation and Behavior Assessment), a system designed to monitor traffic flow, predict parking availability, and flag congestion risks before they escalate.
Such projects rely on close collaboration with stakeholders managing complex environments, and models need to work not only in clean demos, but also in day-to-day operations, where they are exposed to changing conditions and high societal expectations.
ALIVE’s current priorities are to deepen our collaboration with industry and enable us to test our research beyond the laboratory. Industry partners can help provide operational environments, domain expertise, data pipelines, and deployment paths so that systems can be evaluated against practical requirements such as speed, privacy and security protection, hardware constraints, and reliability in a variety of real-world situations. These collaborations also help research teams identify what really matters to end users, helping them turn novel laboratory experiments into life-changing innovations.
For partnership discussions or interview requests, please contact Dr. Patricia Angela Abu at: pabu@ateneo.edu. For any other inquiries, please contact us by email. media.research@ateneo.edu. visit Archeum.Ateneo.edu Learn more about Ateneo’s latest research and innovations.
