DBDebwashis Borman

ESRGAN-MultiBand: Sentinel-2 Super-Resolution

Super-resolution for multiband satellite imagery — adapted ESRGAN with domain-specific perceptual loss.

Deep LearningComputer Vision
View on GitHub

Overview

Adapted the ESRGAN architecture for multiband Sentinel-2 imagery and customized perceptual loss with domain-specific feature extractors. Benchmarked with PSNR, SSIM and LPIPS against several baselines.

Key Highlights

  • Multiband (Sentinel-2) input adaptation for ESRGAN
  • Custom perceptual loss with domain feature extractors
  • Benchmarked on PSNR / SSIM / LPIPS

Tech Stack

PythonPyTorchRemote sensingGANs

Other Projects

All Projects

Projection Mapping in Unity3D

Data Engineering

Spatial calibration + blending for multi-projector image/video content, in Unity3D.

View Project

Arc-Mode Classification + PyQt GUI

Deep LearningComputer VisionPython

Deep-learning pipeline that classifies arc appearance modes from high-speed video, with a PyQt analysis GUI.

View Project
datalanguagemodelnlptokenstextvectorvocab

Multitask NLP Classifier with BERT

NLPDeep Learning

Multitask BERT for paraphrase detection, semantic textual similarity and sentiment analysis.

View Project

Stay Updated

Subscribe to get the latest blog posts, project updates, and data science insights straight to your inbox.

No spam. Unsubscribe anytime.