
Computer Vision & AI Researcher
Heri Septian Adi Nugroho
“A camera and artificial intelligence share one major trait. Both have the absolute power to decide who gets seen, and who can be ignored.”
About
Originally from Jakarta, Ian spent years working in the visual production industry. He built video campaigns for global platforms and saw firsthand how a single image could reshape the sentiment of millions. But as he began documenting educational inequality across various islands, he realized a fundamental problem.
Studio-polished visuals often erased the raw, messy reality of the very communities they tried to represent. This unease pushed him to leave commercial production and shift his focus to how technology processes visual elements (visual epistemology).
Today, Ian delves into computer vision models. He focuses on how algorithms translate and generate educational illustrations. If left uncontrolled, generative AI risks wiping out the rich cultural nuances of 3T schools, replacing them with standardized, heavily biased imagery.
As sons and daughters of the nation, we recognize that visual media is the primary avenue for absorbing knowledge in areas without lab facilities. Because we care for 3T students, Ian ensures our systems read their native environments correctly. We believe that applying AI for education and AI for accurate equity means building technology that can see Indonesian students essentially as they are, without erasing their cultural context.
Activity Gallery
Visual intelligence research and modeling
Videography and field documentation
Analyzing AI-generated technical visualizations
Engineering human-centered visual systems
Vision
Creating an AI-driven visual ecosystem that does not alter rural reality, but honestly projects the region's natural and cultural richness.
Mission
Develop AI systems capable of understanding local visual environments. Prevent the use of image generators that erase regional traditions. Build visual technologies that genuinely aid the learning process in low-resource areas.
Why AFIRMASI?
AFIRMASI does not treat AI as a neutral bystander. In regions without books and with limited equipment, images become the main learning source. We design technical safety boundaries so that machine-generated images do not provide a false context.