Mayank Mahajan

 Developed software for creating artificial data, which is used for training robust deep neural networks.  Distance calculation object tracking for detected objects.  Increased runtime of object detection application by implementing openCV pre-processing on GPU with the help of custom CUDA kernels.  Calibrating neural network in order to save processing time by almost 50% on NVIDIA platforms.  Algorithm development for eliminating false positives after neural network processing.

a) Developed GUI for sensor installation based on algorithms developed for GPS. b) QR code generation, wireless printing support, updating current position on OSM map was added on the GUI application. c) Solved problems with detection setup for better detection accuracy. d) Path planning for gluing with Mitsubishi RV-3SB for different sensors. e) Developed algorithm for detecting unusual behavior on detection accuracy with help of machine learning.

Ich mache mein Master Abschlussarbeit und ich muss eine prototyp mit IMU sensor für Wanderkennung zu entwicklen.

I developed algorithm for object detection with Cascade classifier. The objective of the project was to use machine learning for learning an object and classify that particular object on a live stream. The distance of the robot and the object was calculated using the single camera.

I with a team, was responsible to detect any defects (if any) which are there in the car after assembly of the car is done.

Image Processing, Sensor Fusion, CFD simulations and CAD designs