🎯 Multimodal DSP Challenges

An advanced suite for image restoration, audio denoising, and phonetic formant analysis using Python 3.14.

Author: Mobi

---

✨ Overview

This project showcases cutting-edge Digital Signal Processing (DSP) techniques applied across multiple modalities including image processing and audio analysis. Built with high-performance Python 3.14, it demonstrates sophisticated algorithms for noise removal, spectral analysis, and phonetic characterization.

📋 Table of Contents

• Features
• Installation
• Quick Start
• Project Structure
• Challenges
• Results
• Dependencies
• License

🚀 Features

• 🖼️ Image Processing - Frequency-domain noise removal and blur estimation
• 🔊 Audio Analysis - IIR filtering, spectral leakage, and formant extraction
• 📊 Spectral Visualization - Comprehensive plots and analysis reports
• 🌐 Interactive Report - Live Persian dashboard with real-time simulations
• ⚡ High Performance - Optimized with NumPy and SciPy
• 🎯 Modular Design - Clean, reusable components for each task

📦 Installation

Prerequisites

• Python 3.12 or higher
• uv package manager

Setup

# Clone the repository
git clone https://github.com/Mobiwn/multimodal-dsp-challenges.git
cd multimodal-dsp-challenges

# Install dependencies
uv sync

🏃 Quick Start

Run the challenges individually to generate analysis results in the results/ folder:

# Image Processing Tasks
uv run src/image_dsp.py

# Audio Processing Tasks
uv run src/audio_dsp.py

# Spectral Analysis
uv run src/spectral_utils.py

🌐 Interactive Report

Explore the DSP challenges interactively with a Persian-language dashboard hosted on GitHub Pages.

• Live Demo: View Interactive Report
• Features include real-time notch filter adjustments, motion blur analysis, spectral leakage comparisons, and vowel formant visualizations.

📁 Project Structure

multimodal-dsp-challenges/
├── src/
│   ├── audio_dsp.py       # Audio processing and IIR filtering
│   ├── data_utils.py      # Data loading and utilities
│   ├── image_dsp.py       # Image processing in frequency domain
│   └── spectral_utils.py  # Spectral analysis tools
├── data/                  # Input audio files
│   ├── input_speech.wav
│   └── test_speech.wav
├── docs/
│   └── index.html         # Interactive Persian report (hosted on GitHub Pages)
├── results/
│   ├── audio/            # Processed audio output
│   └── figures/          # Analysis visualizations
├── pyproject.toml        # Project configuration
├── requirements.txt      # Python dependencies
└── README.md            # This file

🎮 Challenges

Challenge 1: Notch Filtering in Image Processing 🖼️

Goal: Remove periodic noise from images using 2D frequency domain filtering.

Techniques:

• 2D FFT transformation
• Frequency domain notch filter design
• Inverse FFT reconstruction

Challenge 2: Motion Blur Estimation 📸

Goal: Estimate blur length in motion-blurred images using spectral nulls.

Techniques:

• Spectral null detection
• Radial averaging in frequency domain
• Blur parameter estimation

Challenge 3: IIR Notch Filter 🔊

Goal: Design and implement IIR notch filter for 50Hz hum removal.

Techniques:

• IIR filter design
• Pole-zero plot analysis
• Stability assessment
• Real-time audio filtering

Challenge 4: Spectral Leakage Analysis 📊

Goal: Understand and visualize spectral leakage effects with different windows.

Techniques:

• Windowing functions (Hann, Hamming, Blackman)
• FFT bin analysis
• Leakage quantification

Challenge 5: Vowel Formant Extraction 🗣️

Goal: Extract and analyze vowel formants using LPC modeling.

Techniques:

• Linear Predictive Coding (LPC)
• Formant frequency extraction
• Phonetic vowel characterization

📈 Results

All analysis results are automatically generated and saved to:

• Audio Outputs: results/audio/ - Cleaned and processed audio files
• Visualizations: results/figures/ - High-quality analysis plots

🛠 Dependencies

Core dependencies managed by uv:

Package	Version	Purpose
NumPy	≥2.4.0	Numerical computing
SciPy	≥1.16.3	Signal processing
Matplotlib	≥3.10.8	Data visualization
OpenCV	≥4.11.0	Image processing
scikit-image	≥0.26.0	Image algorithms
SoundFile	≥0.13.1	Audio I/O
Streamlit	≥1.52.2	Web interface

See pyproject.toml for complete dependency list.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

👨‍💻 Author

Created with ❤️ by Mobi

🤝 Contributing

Contributions, issues, and feature requests are welcome! Feel free to check the issues page.

📜 Acknowledgments

• Built with modern Python 3.14 features
• Powered by scientific Python ecosystem
• Inspired by advanced DSP challenges

---

Made with 💙 by Mobi | GitHub