Nabanita Das

Data analysis and visualization. Advice customers regarding errors in data collection. Research and development(data-driven) of machine learning algorithms for prediction of wear and tear of tools used in the production system by düspohl Maschinenbau GmbH.

Topic: Prediction of Geometrical Features from technical product specifications Tasks: Data analysis and visualization. Research and development of unsupervised machine learning techniques for data annotation. Research and development of supervised machine learning techniques for classification.

Development of a Framework for Heterogeneous High-Performance Image Processing Applications. Modified existing software to correct errors and optimize efficiency.

Research and development of Computer Vision algorithms, eg. lane detection, optical character recognition(OCR). Research on machine learning algorithms and implementation of state-of-the-art algorithms. Automation of processes for calibration of Fluke devices. Development of techniques for medical image enhancement for international clients. Customer interaction for requirements analysis

Focus: Signal, Image, and Speech Processing Projects: Real-time 3D object detection and pose estimation, development of stereo camera using webcams, Image classification, Automatic speech recognition (Natural Language Processing), Development of a multi-class classification model with Random Forest, Neural Networks, and SVM. Relevant courses: Statistical signal processing, Statistical machine learning, Digital image processing, Robotics, etc.