VIDEN

Video Enhancer | Python 3 / FFmpeg / ffprobe / NCNN / Vulkan

A local Python CLI for enhancing video with Real-ESRGAN. It extracts frames, optionally denoises them, upscales each frame on the GPU, then rebuilds the video and restores its original audio.

Tile-based processing helps manage GPU memory, while ffprobe keeps the output dimensions and frame rate aligned with the source.

How it works

Frame extraction

Read dimensions and frame rate with ffprobe, then decode the video into numbered RGB frames. Optional working dimensions preserve aspect ratio.

Upscaling comparison

Input / bicubicReal-ESRGAN
Examples from the original Real-ESRGAN project.

Inside the pipeline

GPU inference
A subprocess runs the NCNN backend with local .param and .bin model files. Model selection, scale, and tile size are passed explicitly.
Memory control
Tiled processing limits GPU memory use. The CLI accepts 2× and 4× scaling; the selected model must support the requested scale.
Video handling
ffprobe reads dimensions and frame rate. Working dimensions preserve aspect ratio and are rounded to even values for video encoding.
Progress & cleanup
Progress counts completed PNG frames. Backend failures retain a log during processing; temporary working files are cleaned up when the run finishes.
Audio
The pipeline extracts the original audio when present and muxes it back into the reconstructed video.

Model & quality tradeoffs

Model scale

Scale depends on the selected NCNN model. The current default is RealESRGAN_General_WDN_x4_v3.

Tile size

Smaller tiles use less GPU memory. Denoising can reduce noise, but may also smooth useful texture.

Temporal detail

Frames are enhanced independently, so flicker and changing textures should be checked across the finished video.

VIDEN handles the orchestration and video workflow; the restoration model and NCNN/Vulkan runtime come from upstream projects. Generated detail is an estimate, not a guarantee of original content.