According to Decrypt, Resemble AI, a startup focused on audio fidelity, has introduced 'Resemble Enhance,' an open-source tool designed to significantly improve the quality of historical audio. The service can take distorted recordings of historical speeches and apply AI to make them sound like they were recorded or broadcast recently. Resemble Enhance combines a sophisticated denoiser, which removes static background noise, with an AI-powered speech enhancer to not only eliminate unwanted noise but also enrich the overall audio quality.
The Resemble AI denoiser uses UNet, an AI model that helps separate different types of sounds in a recording, filtering out unwanted noise and leaving just the speech. After UNet's work, the enhancer module extends audio bandwidth and corrects distortions, ensuring the final output is noise-free and possesses the richness of contemporary recordings. As an open-source tool, Resemble Enhance is accessible at no cost, making it a compelling option in the traditionally expensive market of media restoration services.
Industries reliant on clear audio quality, such as podcasting, entertainment, and education, are the primary beneficiaries of Resemble Enhance. The tool also offers a new lease on life to historical recordings, potentially providing clearer insights into the past. The open-source nature of Resemble Enhance positions it as a potentially disruptive force in a market currently dominated by high-cost proprietary solutions. Users can experiment with Resemble Enhance by visiting the official Resemble AI website or downloading their models from the project's official Github page.
