At IIIT-Bangalore, our students continue to prove that research knows no boundaries. One such inspiring story is that of Sayak Chowdhury (MS2022), who recently represented the institute on a global stage.
In the second week of September, Sayak presented his research at the 34th International Conference on Artificial Neural Networks (ICANN), hosted at Vytautas Magnus University, Kaunas, Lithuania. ICANN is one of Europe’s leading conferences in neural networks and AI, and presenting there was a significant milestone in Sayak’s research journey.
His paper, “Robustness Verification for Object Detection using Set-Based Reachability Analysis”, co-authored with Hardik Khandelwal (IMT2020) and Prof. Meenakshi D’Souza, tackles a problem that’s vital in today’s AI-driven world. Imagine autonomous cars or security camera, what if their AI made a mistake because of unexpected situations? Sayak’s research ensures these systems can handle uncertainty safely and reliably. Using a set-based reachability analysis, his team developed a way to mathematically check that AI predictions remain consistent, even when faced with tricky inputs. In short, it’s about making AI smarter, safer, and more trustworthy.
Within the three-member team, Sayak took the lead at ICANN, standing before an international audience as the sole presenter. Delivering a 15-minute oral presentation in the Senate Hall-broadcast live on YouTube-he brought their research to life with clarity and confidence. The pressure was immense, but so was the pride.
Beyond the talk, the conference was five days of inspiration. Sayak met researchers from across Europe, exchanged ideas, learned different perspectives, and even soaked in the local culture-beautiful churches, folk dances, and a lively gala dinner organized by ENNS. Lithuania, with its warm people and welcoming atmosphere, left a lasting impression. This was not just his first international presentation, but also a moment of representation showcasing the spirit of IIIT-Bangalore’s research community on a global platform. His journey proves that research can feel like exploring a forest-you don’t always know where you’ll end up, but every new path opens the door to something extraordinary.