Upload 3 files

README.md

@@ -0,0 +1,188 @@
+# Infant Cry Classifier
+
+## Table of Contents
+1. [Introduction](#introduction)
+2. [Problem Statement](#problem-statement)
+3. [Solution](#solution)
+4. [Importance and Need](#importance-and-need)
+5. [How It Works](#how-it-works)
+6. [How to Use This Model](#how-to-use-this-model)
+7. [Model Performance](#model-performance)
+8. [Limitations and Considerations](#limitations-and-considerations)
+9. [Future Developments](#future-developments)
+10. [Contributing](#contributing)
+11. [License](#license)
+12. [Acknowledgments](#acknowledgments)
+
+## Introduction
+
+The Infant Cry Classifier is an advanced machine learning model designed to analyze and categorize different types of infant cries. This innovative tool aims to assist parents, caregivers, and healthcare professionals in understanding and responding to infants' needs more effectively.
+
+## Problem Statement
+
+Interpreting an infant's cries can be challenging, especially for new parents or in high-stress situations. Babies communicate their needs primarily through crying, but distinguishing between different types of cries (e.g., hunger, discomfort, tiredness) can be difficult. This uncertainty can lead to:
+
+1. Increased stress for parents and caregivers
+2. Delayed response to the infant's needs
+3. Potential misinterpretation of the baby's requirements
+
+## Solution
+
+Our Infant Cry Classifier addresses these challenges by:
+
+1. Analyzing audio recordings of infant cries
+2. Extracting relevant acoustic features
+3. Classifying the cry into predefined categories (e.g., belly pain, burping, discomfort, hunger, tiredness)
+
+## Importance and Need
+
+### 1. Enhanced Infant Care
+
+By accurately identifying the reason behind an infant's cry, caregivers can respond more promptly and appropriately to the baby's needs. This can lead to:
+
+- Improved infant comfort and well-being
+- Reduced stress for both the infant and caregiver
+- Better overall care and nurturing
+
+### 2. Medical Applications
+
+In healthcare settings, the Infant Cry Classifier can be a useful diagnostic tool:
+
+- Assisting pediatricians in identifying potential health issues
+- Supporting early detection of certain conditions that may affect an infant's cry patterns
+- Providing objective data to complement clinical observations
+
+### 3. Research Opportunities
+
+This model opens up new avenues for research in:
+
+- Infant communication and development
+- Early childhood psychology
+- Acoustic analysis of infant vocalizations
+
+## How It Works
+
+1. **Data Collection**: The model is trained on infant cry audio samples, carefully labeled with their corresponding causes.
+
+2. **Feature Extraction**: Advanced signal processing techniques are used to extract relevant acoustic features from the audio samples.
+
+3. **Machine Learning**: A sophisticated machine learning algorithm is employed to learn the patterns associated with different types of cries.
+
+4. **Classification**: When presented with a new audio sample, the model analyzes it and classifies it into one of the predefined categories.
+
+## How to Use This Model
+
+### Prerequisites
+
+- Python 3.7 or higher
+- Required libraries (install via pip):
+  ```
+  pip install numpy pandas scikit-learn joblib librosa
+  ```
+
+### Installation
+
+1. Clone this repository:
+   ```
+   git clone https://github.com/your-username/infant-cry-classifier.git
+   cd infant-cry-classifier
+   ```
+
+2. Download the pre-trained model files:
+   - `model.joblib`: The trained SVM model
+   - `label.joblib`: The label encoder used for cry categories
+
+### Usage
+
+1. Import the necessary libraries:
+
+```python
+import joblib
+import librosa
+import numpy as np
+```
+
+2. Load the pre-trained model and label encoder:
+
+```python
+model = joblib.load('model.joblib')
+le = joblib.load('label.joblib')
+```
+
+3. Define the feature extraction function (make sure this matches the function used during training):
+
+```python
+def extract_features(file_path):
+    try:
+        # Load audio file and extract features
+        y, sr = librosa.load(file_path, sr=16000)
+        mfcc = np.mean(librosa.feature.mfcc(y=y, sr=sr, n_mfcc=40,n_fft=n_fft,hop_length=hop_length,win_length=win_length,window=window).T,axis=0)
+        mel = np.mean(librosa.feature.melspectrogram(y=y, sr=sr,n_fft=n_fft, hop_length=hop_length, win_length=win_length, window='hann',n_mels=n_mels).T,axis=0)
+        stft = np.abs(librosa.stft(y))
+        chroma = np.mean(librosa.feature.chroma_stft(S=stft, y=y, sr=sr).T,axis=0)
+        contrast = np.mean(librosa.feature.spectral_contrast(S=stft, y=y, sr=sr,n_fft=n_fft,
+                                                      hop_length=hop_length, win_length=win_length,
+                                                      n_bands=n_bands, fmin=fmin).T,axis=0)
+        tonnetz =np.mean(librosa.feature.tonnetz(y=y, sr=sr).T,axis=0)
+        features = np.concatenate((mfcc, chroma, mel, contrast, tonnetz))
+        # print(shape(features))
+        return features
+    except:
+        print("Error: Exception occurred in feature extraction")
+        return None
+```
+
+4. Use the model to classify a new cry audio:
+
+```python
+def predict_cry(file_path):
+    # Load the saved model and LabelEncoder
+    loaded_model = joblib.load('model.joblib')
+    loaded_le = joblib.load('label.joblib')
+    
+    # Extract features from the new audio file
+    features = extract_features(file_path)
+    
+    if features is not None:
+        # Reshape features to match the input shape expected by the model
+        features = features.reshape(1, -1)
+        
+        # Make prediction
+        prediction = loaded_model.predict(features)
+        
+        # Convert prediction back to original label
+        predicted_label = loaded_le.inverse_transform(prediction)
+        
+        return predicted_label[0]
+    else:
+        return "Error: Could not extract features from the audio file"
+
+# Example usage
+file_path = 'path/to/your/file.wav'
+result = predict_cry(file_path)
+print(f"Predicted cry type: {result}")
+```
+
+### Integration
+
+You can integrate this model into your own applications, such as:
+
+- A mobile app for parents
+- A monitoring system for nurseries
+- A research tool for pediatric studies
+
+## License
+
+This project is licensed under the MIT License - see the `LICENSE.md` file for details.
+
+## Acknowledgments
+
+- Thanks to all the parents and caregivers who contributed audio samples
+- Pediatric researchers who provided domain expertise
+- Open-source community for various tools and libraries used in this project
+
+## Conclusion
+
+The Infant Cry Classifier represents a significant step forward in leveraging technology to support infant care. By providing insights into an infant's needs through cry analysis, we aim to enhance the quality of care, reduce parental stress, and contribute to the overall well-being of infants and their caregivers.
+
+We welcome feedback and contributions from the community to further improve and expand this important tool. Together, we can make a meaningful difference in the lives of infants and their families.

label.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a82740f73e162c78aa43903814af40bae174b3c9a84c1b823f282089e0ff9791
+size 527

model.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c640cd82970f11fdd3db4147251b8fe7e28112968c1eca06532e1fca3c0aab2
+size 113695