I’m Vikas Desai, a deep learning researcher with interests in developing limited supervision techniques based on active learning and meta learning for label intensive computer vision tasks such as object detection and semantic segmentation. I’m fortunate to have been a part of Dr. Vineeth N Balasubramanian’s research group - Lab1055 at IIT Hyderabad, while closely collaborating with Dr. Wei Guo at Field Phenomics Lab, University of Tokyo. During my time at IITH, my research focused on using active learning techniques for object detection with applications to precision agriculture and plant phenotyping. I have done internships at University of Tokyo (Summer 2018) and AIST Tokyo (Summer 2019). I currently work at Qualcomm in deep learning automation for mobile platforms.
- NEW! (Jul 2020) Our tool EasyRFP for deep learning based plant phenotyping using Edge Computing has been accepted in (1) ECCV 2020 Demos and (2) CVPPP Workshop, ECCV 2020.
- NEW! (Apr 2020) Our work on fine grained sampling for Active Learning in Object Detection has been accepted in VL3 Workshop, CVPR 2020.
- (Mar 2020) Our survey article on Deep Learning in Agriculture appears in the March edition of ACCS Journal India.
- (Mar 2020) Paper on point supervision based active learning for object detection in cereal crops, got published in Plant Methods, 2020 Edition (Impact Factor: 4.4).
- (Sep 2019) Attended BMVC ‘19 to present our paper on adaptive supervision for object detection.
- (Jul 2019) Paper on computer vision based rice panicle detection got published in Plant Methods, 2019 Edition (Impact Factor: 4.4)
- (Jul 2019) Our paper on adaptive supervision for object detection got accepted in BMVC ‘19, Cardiff, UK.
- (Jun 2019) Started summer internship at AIST, Tokyo.
- (May 2018) Started summer internship at Field Phenomics Lab, UTokyo.
- (Jul 2017) Joined IIT Hyderabad as a Masters Student, in Computer Science.
- (Jun 2017) Graduated in Electronics & Communication Engineering from JNTU Hyderabad.