About this project

QualityScaler is a Windows application that uses AI to enhance, upscale, and de-noise photographs and videos. It features an elegant GUI, supports multiple GPUs, and handles various image (jpg, png, tif, bmp, webp, heic) and video (mp4, webm, mkv, flv, gif, avi, mov, mpg, qt, 3gp) formats. Key features include automatic image tiling to avoid VRAM limits, resizing before upscaling, interpolation between original and upscaled files, and video upscaling with stop/resume. Privacy-focused, it runs entirely offline. Built with Python using PyTorch, ONNX, OpenCV, and CustomTkinter, it requires Windows 10/11, 8GB RAM, and a DirectX12-compatible GPU with 4GB VRAM. The roadmap includes hardware-accelerated encoding, multi-threading improvements, and new AI models. Users can run it from source with Python and VSCode, or download pre-built releases.