While many AI companies are racing to find ways to use this technology to enhance or even create video, Resemble AI focuses on audio fidelity. The startup, which provides an AI voice generator that businesses can use to create realistic human-like voiceovers, has launched ‘Resemble Enhance’, an open source tool designed to drastically upgrade the quality of past audio.
A new service can take distorted and ambiguous recordings of long-forgotten historical speeches and then apply AI to make them sound as if they were recorded or broadcast yesterday.
According to the Canadian company, Resemble Enhance distinguishes itself with its dual-module approach that combines AI-based voice enhancement with a sophisticated noise-canceling device that removes static background noise and hiss. This combination not only eliminates unwanted noise but also improves the overall quality of the audio.
Even though there are other audio restoration products on the market, Resemble’s combination of technologies can be a meaningful differentiator.
How does it work?
Resemble AI denoiser uses UNet, an AI model that helps separate the different types of sounds that appear in your recordings. It’s great for filtering out unwanted noise from your audio track, focusing only on your voice.
Once UNet has done its work, the enhancement module kicks in to expand the audio bandwidth and correct distortion. This dual functionality ensures that the final output is not only noise-free, but also rich in modern recordings, the company says.
As an open source tool, Resemble Enhance is free to access, making it an attractive option in the traditionally expensive media restoration services market. Key beneficiaries of Resemble Enhance are industries that rely on crystal-clear audio quality, including podcasting, entertainment, and education. This tool also brings new life to historical records, potentially providing clearer insight into the past.
The launch of this tool comes at a time when demand for high-quality digital content is at an all-time high. Meanwhile, Resemble Enhance’s open source nature positions it as a potentially disruptive force in a market currently dominated by expensive, proprietary solutions.
The convergence of AI in audio and video enhancement will lay the foundation for more comprehensive media restoration solutions.
By combining this tool with other video enhancers (such as GPEN or the well-known GFPGan) that use generative AI or other models to enlarge and enhance images and faces, users can now achieve professional results on their own computer with minimal investment. You can get it.
To experiment with Resemble Enhance, users can visit the official Resemble AI website or download the model from the project’s official Github page.
Edited by Ryan Ozawa.