Aeyez is an egocentric spatial reasoning and safety intelligence platform for first-person visual environments. It focuses on scene understanding, temporal reasoning, hazard detection, navigation assistance, and accessibility-oriented visual guidance.
Designed multimodal reasoning workflows for first-person video and dynamic environments.
Supported benchmark and evaluation pipelines for spatial intelligence tasks.
Worked on temporal scene interpretation and embodied agent interaction logic.
A low-cost outdoor shelter concept for feeding and monitoring stray cats in community spaces. The system combined lightweight computer vision, embedded sensing, and automated feeding mechanisms.
Explored cat detection, feeder triggering, low-power deployment, and weather-resistant housing.
Designed around practical community welfare rather than a lab-only prototype.
computational social systems
Short-Form Media Ecosystems
NeurIPS workshop 2026 pending :)
A research prototype studying how short-form media trends mutate, compete, and adapt under algorithmic selection pressure across platforms such as TikTok and Instagram Reels.
climate intelligence
CAFE / CredibleControls
NeurIPS workshop 2026 pending :)
A climate adaptation and forest dynamics modeling direction focused on forest-risk forecasting, ecological counterfactuals, and evidence-grounded planning under environmental uncertainty.