Research Projects

This project will develop a suite of advanced computer vision technologies, including algorithms and applications, focusing on road, traffic and road user behavior understanding in developing countries such as India. As the developed Western world focuses on self-driving vehicles, the opportunity has surfaced for Indian researchers to develop solutions for lower levels of autonomy that can advance road mobility and safety in the unstructured conditions of Indian roads. Examples of such technologies that will be developed in this project include road surface and quality inspection (eg. pothole detection, water stagnation detection, automatic road quality audit), traffic congestion detection, driver alertness detection, and related computer vision modules that can enhance the safety on Indian roads. In addition to the development of a suite of applications, this project will result in the creation of large data corpuses for enabling research and development in road mobility in unstructured conditions, which will be shared with the wider research community. The availability of such data corpuses will also allow us to mine patterns of traffic (e.g. mean distance between vehicles in a particular area) as well as user behavior (e.g. number of sudden brakes or turns), which could be useful for traffic management by law enforcement and other government agencies. The project team brings together researchers from IIT-Hyderabad, IIIT-Hyderabad and IIT-Delhi with extensive experience in computer vision and machine learning, as well as their applications in real-world systems including transportation (team background and selected recent efforts are enclosed with this proposal). IIIT-Hyderabad also hosts an Applied Research Lab which focuses on translating research outcomes to usable applications, which will be pivotal for the success of this project.