Ankit Satpute

- Algorithms related to Adaptive Computer Vision Systems. - Application to Robotics (Planning based on the image data, Object tracking and automatic path prediction based on Machine Learning Methods ). - Fully CNN for fast target detection. - Design and Analysis of Swarm robotics algorithms in the area of Autonomous Robotics. - Proposed model for cloud-based instant training to avoid onboard computational calculation in case of autonomous Aerial vehicle target detection and tracking using CNN.

- Implemented a data-driven model for person tracking - Evaluated Conventional Image Processing methods for human tracking within the overlapping as well as non-overlapping camera - A novel method for Person re-identification based on physical appearance and posture changes. - Made the proposed model robust to viewpoint illumination, occlusions and trajectory loss. - Applying the proposed methods to ground-based as well as aerial based systems.

- Detection of Abnormalities in the retina using Deep Learning. - FasterRNN_NAS and InceptionV3 CNN Deep Neural Network Architecture are used for the detection and classification of vital parts from the human eye. - Dataset: Kaggle Retinal Image dataset for Diabetic Retinopathy.

- Study of Laser Designated devices and types of equipment. - Design of Digitally controlled Power Board for generating a coded pulse from Laser Diode Array. - Countermeasures techniques useful in on-field war system.

- Pattern Recognition - Artificial Neural Networks - Digital Image Processing - Artificial Intelligence