spectral-analysis

Rust implementation of the Resonate algorithm for low-latency spectral analysis, with Python and WebAssembly bindings

Detect fake lossless audio (transcoded from lossy) via FFT spectral analysis. Fast, batch, written in Rust.

A Python package for interacting with aus-rust's analysis capability

Detects guitar chords from raw audio using FFT-based spectral analysis and a modular Rust scoring engine. Includes batch tools, spectrograms, and extendable chord templates.

Real-time audio streaming system using RTP transport (RFC 3550) and Opus encoding (RFC 6716). This project demonstrates end-to-end systems architecture from audio capture through network transport to playback, with focus on production concerns: network resilience, observability, and audio intelligence through ML model integration.

Clean-room healthcare signal analytics for edge, bedside, and remote monitoring. 63 Rust functions, 127 tests, Kalman filter, streaming API, criterion benchmarks. Zero ML dependencies.

Spectral analysis of conversational prosody: Fourier decomposition of agent dialogue timing, harmonic cadence detection

Pure Rust ML primitives — spectral analysis, protein folding, Anderson localization. 4,500+ validation checks. Sovereign ML on commodity GPUs. AGPL-3.0

AI token spending intelligence: wavelet decomposition, conservation law budgeting, and spectral model correlation analysis

Spectral audio analysis in Rust + WASM with SIMD — STFT, CQT, wavelets, pitch detection, source separation, and MFCCs running in the browser in real time.