The project aimed to develop and deploy an autonomous interactive robotic wheelchair in hospital spaces, university campuses, and even residential spaces. The project involves building a robotic wheelchair equipped with sensors like depth cameras, LiDARs and a microphone to effectively interact with the dynamic environment and follow complex commands given by the operators.

We implemented Mapping, Localisation, Navigation, Obstacle Avoidance(Static & Dynamic), Human Following, Human Velocity Prediction etc. on the wheelchair using both LiDar and RGBD camera (one at a time). The project is demonstration dependent as much as it is research dependent. Current research addresses the problem of Referring Image Segmentation, that uses language commands to attend to correct entities in images. Peripheral projects include following a specified person in a crowd, finding a person/scene using scene recognition, etc., and localizing efficiently in long corridors.