This is an Edge Impulse AI Actions block that uses Audio Spectrogram Transformers from HuggingFace to automatically label audio data. You can use this repo as the basis for custom tasks that use big ...

An experimental performance tool that converts images captured from your webcam into sound by interpreting them as frequency spectrograms. One can either click on the button to take an image or ...

Automatic marine animal/ship-radiating sound discrimination plays a vital role in many real-world sound-aware applications such as marine pollution monitoring, acoustic target detection and tracking, ...

Abstract: In this work, we propose CleanMel, a single-channel Mel-spectrogram denoising and dereverberation network for improving both speech quality and automatic speech recognition (ASR) performance ...