Publications & Research Output

AbstractAI-generated synthetic media — multimodal deepfake content — has become a pressing concern, with older adults (55+) and younger users (11-18) at greater risk than the general adult population. This thesis uses participatory design, combining interviews and surveys, to understand how these two communities perceive deepfakes. Building on participants' insights, we trained an awareness model and designed a mobile application that serves as a practical AI-awareness tool. The work is explored through Computer Vision, Human-Computer Interaction, and Social Computing perspectives.

AbstractSIDAS is a wearable AI navigation system designed to help visually impaired people move around more safely and independently. It uses a smartphone camera, AI-based depth perception, and onboard sensors to understand obstacles and decide the safest direction in real time, guiding the user through smartwatch vibrations and spatial audio — while keeping critical safety decisions on-device for low latency. A manuscript describing the system design is in preparation.