Batch denoising for images and audio.
Oath is a command-line application for batch denoising image and audio datasets. It is designed to support repeatable processing workflows for collections of noisy media while remaining simple enough to run on local hardware.
The project focuses on providing a clean and practical interface for running denoising jobs across directories of files. Rather than building isolated scripts for individual datasets, Oath aims to offer a reusable tool that can be applied to a variety of image and audio processing tasks.
Oath is intended to be used primarily as a CLI tool. Jobs can be executed locally through command-line commands, making it easy to integrate with shell scripts, scheduled tasks, or other automation workflows.
The project emphasizes:
- batch processing of image and audio datasets
- simple command-line workflows
- reproducible processing environments
- modular pipelines that can support different denoising methods
While Oath is designed to run well on a single machine, its architecture leaves room for future extensions such as workflow orchestration or distributed execution. The immediate goal, however, is to provide a reliable and straightforward tool for running denoising pipelines on local datasets.