Smart Collar for Pet Dogs: Audio-Based Emotional Analysis

This project focuses on the development of a smart collar for pet dogs, integrating Internet of Things (IoT) and Machine Learning (ML) technologies to monitor pet health effectively. Initially, the collar will perform audio-based emotional analysis, with plans to expand into physiological monitoring as suitable datasets become available.

---

Key Features

Audio Emotion Classification

• Vocalization Monitoring: The collar records audio samples of the dog's vocalizations every 5–10 minutes.
• Emotion Detection: The audio recordings are processed using an ML model trained to classify the dog's emotional state (e.g., happy, angry, sad, etc.).
• Daily Summary: The results are aggregated to provide a daily emotional pattern overview.

---

How to Run the Project

1.Download and Process Audio Data

Next, download the audio files and preprocess them. This will cut them into smaller parts and prepare the data for analysis.

python audioset_download.py

2.Preprocess the Audio Data and Visualize Spectrograms

After downloading the data, run the preprocessing script to visualize some spectrograms:

python audio_preprocessing.py

3. Train the Model

Now, run the appropriate model training script. For example, to train using MobileNetV2, run:

python mobileNetV2.py

4. Test the Trained Model

python model_tester.py

5. Run the Simulation

python simulation.py