👨‍🎓
Somesh Fengade
  • ME
    • 👋About
    • 📚Personal Projects
      • 🌄AI meditations
      • 🤖GPT-Compare
      • 🌙Learning to see in the dark
      • 💸Lending Club analysis
      • ⏬Moodle Downloader
      • 📚Analyzing the complexity of literary passages.
      • 🧪Unit test analysis
      • 🍕Food Vision
      • 📜Skim Lit project
      • 🚋Driver's attention detection
      • ⏳React Countdown Timer app
  • Achievements
    • 🎓Achievements directory
      • 🐣Kaggle Expert (notebook)
      • ❗Amazon Machine learning competition (top 100)
      • 🐶PetFinder.my - Pawpularity Contest ( top 20%)
      • 🐳Happywhale - Whale and Dolphin Identification (top 20%)
      • 🖊️NBME - Score Clinical Patient Notes ( top 50%)
      • 🕸️Tensorflow Barrier Riff challenge(top 42%)
  • Notes and courses
    • 🍂Foundations of LLMs Lesson 1
    • GPT from scratch
    • Introduction to CUDA Python with Numba
      • Custom CUDA kernels in python with Numba
      • Multidimensional grids and shared memory for cuda python with Numba
Powered by GitBook
On this page
  1. ME
  2. Personal Projects

Learning to see in the dark

In this project, we implemented a research paper ( learning to see in the dark ) which mainly converts the night-time images to daytime images.

PreviousGPT-CompareNextLending Club analysis

Last updated 2 years ago

GitHub Link

Live demo link:

Libraries used:

  • Fast-ai

  • Pytorch

  • matplotlib

  • streamlit

  • gdown

The Model was developed on top of UNET architecture. This project can be used to convert night-time images to daytime images in real-time. For the web interface part, we have used Streamlit and it's also deployed on the Streamlit Cloud.

📚
🌙
  • GitHub Link
  • Live demo link:
  • Libraries used:
LogoGitHub - someshfengde/learning_to_see_in_dark: converting night time images to day imagesGitHub
click to open on github
Logohttps://share.streamlit.io/someshfengde/learning_to_see_in_dark/main/app.py
click to open live demo