This is the official repository of the papers "Parameter-Efficient Transfer Learning of Audio Spectrogram Transformers" [IEEE MLSP 2024] and "Efficient Fine-tuning of Audio Spectrogram Transformers ...

WASHINGTON — What began as an inquiry into a mysterious sound in the background of an airplane cockpit voice recording escalated into an unexpected challenge for the nation's top safety investigators.

The National Transportation Safety Board pulled the plug on its entire public docket system on May 21 after discovering that people on the internet had used AI to reconstruct cockpit voice recorder ...

In the latest sign of these AI-heavy times, the National Transportation Safety Board temporarily removed access to its docket system after discovering that voices of pilots who were killed in a UPS ...

Pilots’ voices from the last seconds of a fatal cargo plane crash have been re-created by Internet sleuths using software and AI tools. The spread of reconstructed audio recordings has prompted a US ...

Abstract: Text-to-audio (TTA), which generates audio signals from textual descriptions, has received huge attention in recent years. However, recent works focused on text to monaural audio only. As we ...

ABSTRACT: Advances in AI-based voice production and conversion technologies have made it possible to create deepfake voices that closely resemble real human speech, raising new security challenges in ...

I've been digging into the audio preprocessing in transformers.js and noticed an issue: There are currently no unit tests for the audio_utils module in the JS implementation. The output of spectrogram ...

TL;DR: Musician and acoustic scientist Benn Jordan achieved a potential world first by encoding data onto a live European starling. Using a spectral synthesizer, he transformed an image into sound, ...

Digging deeper into AI "voices" I previously tested all 30 voices provided by Gemini's TTS and categorized each voice based on impression axes such as "high/low pitch" and "feminine/masculine".