B.Tech Electronics & Communication Engineering · Amrita Vishwa Vidyapeetham, Coimbatore · Batch S4-ECE-B1
- 🔭 Currently building a SAR-guided Deep Learning Framework for Multispectral Super-Resolution & 3D Terrain Reconstruction — using CNNs/Transformers to fuse SAR backscatter with optical reflectance, InSAR-derived DEMs, and produce spatially enhanced 3D multispectral outputs. And working on a method to covert the HDF5 file format to a .SAFE format for processing NISAR data on ESA SNAP tool.
- 🌱 Actively learning Linux, Machine Learning, Image Processing, VLSI, Communications System & GATE prep
- 🎯 Long-term goal: GATE → M.Tech (IIT/IISc, Remote Sensing / Radar) → ISRO
- 👯 Looking to collaborate on Signal Processing and Machine Learning projects
- 🤝 Seeking help with CNN/Transformer-based multimodal deep learning in MATLAB
- 📍 Based in Bengaluru, India
- 📫 Reach me at gadi37681@gmail.com
- ⚡ Fun fact: Decent badminton & soccer player. Also a PC gamer 😁
|
SAR + MSI Fusion Pipeline A SAR-guided deep learning framework to super-resolve multispectral imagery using CNNs or Transformers. Learns cross-domain correlations between radar backscatter and optical reflectance. Uses InSAR-derived DEMs for 3D terrain reconstruction. Pipeline: SAR Preprocessing → MSI Preprocessing → Coregistration → InSAR/DEM Generation → Feature Extraction → DL Fusion + Super-Resolution → 3D Reconstruction → Downstream Classification → Validation Applications: Urban mapping · Infrastructure monitoring · Land-cover classification · Post-disaster assessment |
 |
 |
 |
 |
 |
 |
"The cosmos is within us. We are made of star-stuff." — Carl Sagan