🧠 HushAI – AI-Enhanced Adaptive Noise Cancellation System
"Silence is golden, but not always free. HushAI makes it intelligent...!!!!!"
🎯 Project Overview HushAI is an AI-enhanced, software-based adaptive noise cancellation system built entirely in Python. It simulates how advanced noise cancellation works—using adaptive LMS filtering—while combining it with AI-generated voice input and a modern dashboard to visualize and compare clean vs noisy vs denoised audio.
This project is perfect for showcasing skills in:
✅ Python & Signal Processing.
✅ Real-Time Audio Filtering.
✅ AI + TTS Integration.
✅ Data Visualization.
✅ Interactive Dashboards with Streamlit for live Graph Plotting.
⚙️ Tech Stack
- Language Python3.
- Audio Processing NumPy, SciPy, Librosa, SoundFile.
- Visualization Matplotlib.
- TTS Google Text-to-Speech (gTTS).
- Dashboard StreamLit.
- Dataset UrbanSound8k (for Background Noise).
🚀 Features 🎙️ TTS-based clean voice generation.
🌪️ Realistic noise mixing with UrbanSound8K samples.
🧠 Adaptive Noise Cancellation using LMS filtering.
🎧 Before vs After waveform plotting.
🖥️ Streamlit Dashboard to visually/audio-wise compare outputs.
💾 Lightweight, hardware-free implementation.
🧠 Theory Behind LMS Filtering The Least Mean Squares (LMS) algorithm is a classic adaptive filter that continuously adjusts its parameters to minimize the difference between a desired and an actual signal. It is commonly used in noise cancellation headsets and echo suppression systems.
In HushAI, LMS is used to estimate and subtract the noise from the mixed signal based on a reference noise input.
<Thank You Kindly....!!!!>
Screenshots ->
1.
DashBoard 👆👆👆
Clean Voice Waveform 👆👆👆
Noisy Waveform 👆👆👆
Noise Cancelled Waveform 👆👆👆