Development and validation of deep learning algorithm using ultra-wide-field Optos images for early detection of Diabetic Retinopathy in 45-degree single-field, as Co-PI,

Funded by UNSW School of Optometry and Vision Science: Diabetes Research Grant from the Elizabeth O'Beirne and Robert and Emmy Mather Trust Fund, Ongoing.

Mobile App for assisting Indian farmers in improving yield by automatic identification of plant diseases (Rice and Tomato) [  Funded by DST SYST, Ongoing]  as PI.

India suffers from annual crop loss of many crores of rupees owing to pest and disease infestation, according to a study by Assocham and Yes bank. Different plants are infested by different diseases; while some are caused by fungi, some are caused by bacteria and some by viruses.  This project aims in automatically identifying key diseases in specific plants using mobile application, so that the farmers can be benefitted. The image database pertaining to the Indian subcontinent, would be employed to train a model that can automatically predict plant diseases. Both traditional image processing techniques as well as the recent advances in convolutional neural networks would be employed for this purpose.