HasnainRaz/Fast-SRGAN

A Fast Deep Learning Model to Upsample Low Resolution Videos to High Resolution at 30fps

What it solves

It addresses the challenge of upscaling low-resolution videos to high-resolution in real-time, providing a faster alternative to standard super-resolution methods.

How it works

The project implements the SR-GAN architecture, utilizing pixel shuffle for efficient upsampling to increase processing speed. It includes a generator model that can be trained on datasets like DIV2K or used via pre-trained weights.

Who it’s for

Developers and researchers looking for a high-performance super-resolution tool capable of processing video frames at speeds suitable for real-time applications (e.g., achieving ~30fps for 720p output).

Highlights

  • Real-time performance: Capable of upsampling to 720p at approximately 27-30 FPS on M1 Pro GPU.
  • Efficient Upscaling: Uses pixel shuffle to optimize the upsampling process.
  • Flexible Training: Integrated with Hydra for easy configuration of hyperparameters via the command line.
  • Pre-trained Model: Includes a generator model pre-trained on the DIV2K dataset for immediate use.

Related

  • Project
    SunOner/sunone_aimbot_2

    A C++ based AI aimbot that uses computer vision models to detect targets and automate mouse movement for gaming.

  • Project
    wenbowen123/BundleTrack

    BundleTrack is a real-time 6D pose tracking framework for novel objects that operates without requiring 3D CAD models of the object or its category.

  • Project
    adipandas/multi-object-tracker

    A Python library providing easy-to-use implementations of various multi-object tracking algorithms and OpenCV-based object detectors for video analysis.

  • Project
    hzxie/CityDreamer

    CityDreamer is a research‑grade PyTorch framework that generates unlimited, photorealistic 3‑D cityscapes by combining a layout generator, a background‑stuff generator, and a building instance generator. It includes code for training on large OSM/Google‑Earth datasets, pre‑trained checkpoints, a web demo, and a CLI for rendering videos.