https://github.com/dhanushi2620/aquasignature
Deep learning model using CRNN and MFCC features to classify underwater sounds and detect foreign threats based on acoustic frequency shifts.
https://github.com/dhanushi2620/aquasignature
acoustic-signature ai-for-defense anomaly-detection deep-learning-models keras librosa mfcc spectrogram tensorflow
Last synced: about 1 month ago
JSON representation
Deep learning model using CRNN and MFCC features to classify underwater sounds and detect foreign threats based on acoustic frequency shifts.
- Host: GitHub
- URL: https://github.com/dhanushi2620/aquasignature
- Owner: Dhanushi2620
- Created: 2025-06-26T12:20:52.000Z (12 months ago)
- Default Branch: main
- Last Pushed: 2025-07-11T17:07:22.000Z (11 months ago)
- Last Synced: 2025-07-11T19:11:45.198Z (11 months ago)
- Topics: acoustic-signature, ai-for-defense, anomaly-detection, deep-learning-models, keras, librosa, mfcc, spectrogram, tensorflow
- Language: Jupyter Notebook
- Homepage:
- Size: 1.55 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 🌊 AquaSignature: Underwater Threat Detection using CRNN
- AquaSignature is a deep learning-powered system designed to recognize and classify underwater acoustic signals such as submarines, torpedoes, and marine animals. The core objective is to detect foreign underwater threats that cross a defined acoustic frequency barrier, making it highly relevant for naval defense and surveillance.
## 🎯 Main Objective
- To automatically detect and classify acoustic anomalies — such as submarines or torpedoes from other countries — based on their marine sound signatures. If the barrier of predefined frequency is crossed, the system raises an alert, helping defense forces monitor underwater activities in real time.
## 🔍 Use Case
- This project simulates a naval surveillance system that can:
- Distinguish between natural marine sounds (e.g., dolphins, ships)
- Identify suspicious patterns like torpedo or submarine movement
- Alert when sound frequency breaks the “safe zone” of underwater activity
## 🧠 Features
- Processes .wav files containing marine acoustic data
- Extracts MFCC features and visualizes spectrograms
- Uses a CRNN (CNN + LSTM) architecture for spatial + temporal learning
- Detects when a frequency threshold is crossed
- Classifies sounds as torpedo, ship, dolphin, submarines etc.
## 🛠️ Technologies
- Python
- Librosa (audio feature extraction)
- TensorFlow/Keras (deep learning model)
- Scikit-learn (label processing)
- Matplotlib (visualization)
## 🔐 Real-World Applications
- Naval submarine detection systems
- Underwater mine or torpedo tracking
- Marine research and anomaly detection
- Coastal security monitoring