Pranava Seth

Developed various Machine Learning and Computer Vision algorithms for automation of processes in multiple industrial sectors.

Worked on development of a two-level Event Detection System(EDS) and a catalog for the data that will be downlinked from the Solar Ultraviolet Imaging Telescope(SUIT), which is a Spectral Photo Imager installed in Aditya-L1.

Worked on developing an Open Source Digital Twin for Cancer and on developing 3D Computational models using Wavelet Decomposition, Machine Learning, Computer Vision, and Neural Operators, to automate the segmentation of brain tumors, integrating it with Parametric Classification.

Worked on development of 2D UltraSound Image Segmentation system, using techniques of Supervised Machine Learning, Deep Learning and Computer Vision, integrating it with Clinical Parametric Data.

Utilized the HelioPhysics Knowledge Database for Vector Labels and Mg II FITS Files for Multi-Wavelength data to contribute to developing the initial phase of the Solar Event Detection System. Utilized Convolutional Neural Network (CNN) and Unsupervised Machine Learning techniques.

Worked on Developing a Prototype of an automation system for Garment Measurement. Used the Concept of Landmark Detection to extract key points and visualize them on the inference Data and further designed the algorithm for measuring the size of the garment.

Worked on a project that dealt with predicting the exact location of a promoter in a DNA at the genomic level. Worked on the project's Machine Learning and Digital Signal Processing side.

Developed a computer vision-based Counter Drone System for the Indian Army. Developed and deployed the detection and tracking component of the system on an NVIDIA Jetson Xavier NX. Also Responsible for the Serial Communication component for sending the detected coordinates to the microcontroller in a live transmission.