GitHub

A professional audio signal processing application that provides frequency domain equalization with real-time visualization and AI-powered audio processing capabilities.

Upload and analyze audio signals in the frequency and time-frequency domains Apply customized equalization using interactive frequency sliders Leverage AI models for source separation (human voices, ...
GitHub

Pure JavaScript mel spectrogram, FFT, and audio feature extraction for ML/AI.

Zero dependencies. Runs in browsers (including Web Workers, WebGPU pipelines) and Node.js. NeMo-compatible output validated against ONNX reference models. Built for developers working with speech ...
IEEE

CleanMel: Mel-Spectrogram Enhancement for Improving Both Speech Quality and ASR

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 ...
IEEE

HP-FFT: A General High-Performance FFT Generator Using High-Level Synthesis

Abstract: The Fast Fourier Transform (FFT) is a foundational algorithm widely used in fields like digital signal processing and machine learning. While High-Level Synthesis (HLS) tools have boosted ...