Biography
I am an AI enthusiast immersed in the dynamic world of technology. My journey in artificial intelligence has led me to a captivating intersection where I delve into the realms of Generative AI, 3D GANs, Transformers, and Multi-modal Machine Learning. I’m passionate about pioneering innovation and am dedicated to exploring the limitless potential of these cutting-edge technologies.
- Artificial Intelligence
- Generative AI
- 3D Reconstruction
- Multi Modal ML
-
Bachelor of Technology, 2019-2023
SRM University
Skills
Generative AI| 3D Reconstruction | Neural Rendering | Multi Modal ML | Video Synthesis | Langchain |
Tensorflow - Keras | Pytorch | GANs | Vision Transformers | NERFs | Volume Rendering |
MongoDB (PyMongo) | Firebase (Pyrebase) | MYSQL (SQLite3)
Django |Flask | Streamlit | FastAPI | Postman | HTML | CSS |
Python | C/C++ | Java
Experience
- Researching and developing novel algorithms for the generation of Synthetic Lip-Synced Videos using GANs, Autoencoders and Transformers.
- Using 3D rendering and Cross Attention Transformers to build architectures for high level video generation in multiple languages.
- Worked under Dr. Sebastian Scherer on vision-based aircraft detection and tracking system for detect-and-avoid applications.
- Contributed to the Airborne Object Tracking challenge using a multi-stage pipeline incorporating optical flow and center tracking.
- Trained the pipeline using the Amazon Airborne Detection Challenge dataset and optimized models using TensorRT and TorchJIT for improved performance
- Working on disease prediction using Multi-Modal Cross Attention Transformer Attention and Vision Transformers with Graph Neural Networks.
- Worked on image feature extractor, explored State of the art architectures - SWIN transformer and VIT Transformer as image backbone for Fashion product image retrieval
- Worked closely on the Camera Health Management System by automating video surveillance using Deep Learning techniques
- Implemented facial recognition, object detection, object tracking, and human detection to track CCTV camera behavior.
- Deployed machine learning and deep learning models on hardware for real time monitoring and analysis of camera health