audio-processing-with-python

Audio tagging is the process of inferring descriptive labels from audio clips (Multi label classification task). This repository contains exploratory code/scripts for audio preprocessing and model fitting for the task of audio tagging and its applications.

High Accuracy model which listens to voices and classifies it as Ai or Real

Terminal-based voice logger with real-time playback and optional compression -- built for fast feedback, idea capture, and speech awareness.

An intelligent speech recognition system that combines OpenAI's Whisper for accurate transcription with dual emotion detection models. Analyzes both audio characteristics (tone, pitch, intensity) and textual content to provide comprehensive emotional context alongside transcriptions.

Interactive AI platform for public speaking and impromptu speech practice with multi-metric audio evaluation, real-time feedback, and personalized AI coaching.

AI system that analyzes urban soundscapes to optimize city planning, reduce noise pollution, and enhance acoustic environments using audio deep learning.

This Python script is for a voice interface chatbot named Jervis. It uses OpenAI's GPT-3.5-turbo-instruct model to respond to user input. The chatbot responds by Elevenlabs Voices. Conversation are saved to MongoDB, and MP3 file local and can be emailed if needed.

A simple python script that recursively searches for files and transforms them to mp3, using ffmpeg.

Industrial audio anomaly detection on the MIMII dataset using MFCC-SVM as a baseline and Transformer Autoencoders

Convenience functions for commonly used digital signal processing plots.

This project automates audio processing by removing silence, transcribing speech to text, and storing the output in an SQLite database. It supports multiple audio formats and leverages Google Speech Recognition for high accuracy.

Conversation-Mixer-Tool: A Python utility to merge two audio files (caller and receiver) into a seamless, conversation-like output. Features include speech detection, bandpass filtering, noise reduction, and smooth audio transitions.

DSP-based audio denoising with quantitative evaluation using RMS energy, crest factor, peak-to-RMS metrics, and time–frequency analysis.

A piano transcriber. Give it a song and it writes out a playable two-hand piano score in MusicXML, built on Demucs, CREPE, madmom, and Basic Pitch.

A high-performance Discord music bot built with Python, discord.py, yt-dlp, and FFmpeg featuring cool UI buttons.

See what audio processing actually does — interactive STFT explorer with an A/B spectrogram difference view.

local transcription with speaker separation and paralinguistics. any language, runs offline

A deep learning speech emotion recognition system using CNN-LSTM, MFCC feature extraction, and audio augmentation to analyze emotional patterns and explore their potential for early depression screening.

⚡ Zero-cost AI video dubbing pipeline: Frame-accurate translation, hybrid pitch speaker diarization, 100% background music preservation & Web UI.