car-noise-report

To detect abnormal car noises, this Skill computes RMS energy and STFT from vehicle audio, correlates features with CAN data like RPM and speed, and detects spikes to classify events. It outputs a structured Markdown diagnostic report with visualizations.

Correlating audio features with CAN data for vehicle diagnostics involves computing STFT and RMS energy from audio files, matching them against CAN signals like RPM and speed, and detecting anomalies. This generates per-take visuals and an overview image.

To generate a structured report from vehicle audio and CAN CSV files, process ZIPs or directories containing audio, can.csv, and metadata.json. The workflow computes audio features, detects spikes, classifies events, and outputs a Markdown report with overview images.

Yes, this audio analysis workflow requires FFmpeg for processing webm files, converting them to WAV format for analysis. It also relies on Python libraries like numpy, scipy, pandas, matplotlib, and soundfile for computing features and generating reports.

Vehicle noise analysis supports audio formats like webm and WAV, along with can.csv files and metadata.json. It processes inputs from ZIPs or directories, computing RMS and STFT to detect abnormal noises and output a diagnostic report.

Combining CAN data with audio for car noise detection allows correlation of acoustic events like spikes with vehicle parameters such as RPM and speed. This contextual classification improves diagnostic accuracy and produces a structured maintenance report.