AGE ESTIMATION BASED ON

VOICE USING DEEP NEURAL

• The project focuses on developing a deep neural network-based system for age estimation from human voice
recordings. Accurate age estimation from audio data holds significant potential for a wide range of
applications, including personalized content delivery, voice-based authentication, and targeted marketing.
From each of these files, 24 features are extracted and used to train the neural network.

• The trained neural network reached a test accuracy equals to 94.48%.Ultimately, the developed age
estimation model has the potential to enhance user experiences in various applications, including voice
assistants, customer service, and age-restricted content delivery.

• Additionally, this research contributes to the growing field of voice-based AI technologies and underscores
the importance of ethical and responsible AI development.

• Our project leverages the capabilities of several Python libraries, including Pandas for dataset management
and CSV file handling, Librosa for audio feature extraction, Scikit-learn for machine learning tasks, and
Keras for training artificial neural networks (ANNs). These libraries collectively empower the development
of a robust and accurate age estimation system.
 SCOPE OF PROJECT

• The applications of this system are wide-ranging, from personalized content delivery to voice-based
authentication and targeted marketing. Leveraging Python libraries like Pandas, Librosa, Scikit-learn, and
Keras, the project ensures robust dataset management, feature extraction, and model training.

• Ethical considerations, including privacy concerns and potential biases, are addressed to promote responsible
AI development. Ultimately, this research contributes to the growing field of voice-based AI technologies and
underscores their potential to enhance user experiences across various domains.
 OBJECTIVE
• The project's main objective is to develop a precise age estimation system using deep neural networks based
on human voice recordings.

• This involves extracting essential features from audio data, training a deep neural network to learn age-related
patterns, and achieving a high level of accuracy in age predictions. Ethical considerations, including privacy
concerns and potential biases, are integral to the project.

• The system's applications span various domains, such as voice assistants, customer service, and targeted
marketing, where accurate age estimation can significantly enhance user experiences.

• By leveraging Python libraries and adhering to responsible AI practices, this project contributes to the
advancement of voice-based AI technologies while emphasizing ethical and accurate age estimation.
 EXISTING SYSTEM

• The existing age estimation system with images or facial recognition is a sophisticated application of
computer vision and machine learning that accurately estimates a person's age by analyzing facial attributes.

• This system employs advanced face detection algorithms and deep learning models to extract valuable
features from facial images, which are then used to predict the individual's age.

• With applications spanning security, retail, entertainment, and healthcare, the system addresses challenges
like facial expression variations and lighting conditions while upholding ethical considerations such as data
privacy and consent.

• Its scalability and accuracy make it a valuable tool for tasks ranging from age verification to targeted
marketing and medical diagnostics.
 DISADVANTAGES

• There is a lack of standardization in age estimation methodologies and evaluation metrics, making it
challenging to compare the performance of different systems and models accurately.

• Widespread use of facial recognition technology, including age estimation, faces resistance from segments of
society concerned about civil liberties and surveillance. Achieving social acceptance remains a challenge.

• There is no voice recognition and estimating capabilities in the existing system.

• By leveraging a diverse dataset and a sophisticated neural network model, the system can reliably estimate
age, offering applications in healthcare, entertainment, and authentication systems. It emphasizes ethical
considerations and rigorous evaluation to ensure trustworthy and privacy-conscious age predictions, marking a
significant advancement in voice-based age estimation technology.

• The artificial neural network was developed with the keras library. The dataframe is split into Training (80%),
Test (10%) and Validation set (10%).

• To train the model, five training stages with different batch sizes were used. In order to obtain the best
performances, in this work were tried combinations of different architectures and batch sizes.
 ADVANTAGES
• The system achieves high accuracy in age estimation, thanks to the powerful deep neural network model. This
accuracy enhances the reliability of age predictions across a wide range of applications

• The system represents a significant advancement in voice-based age estimation technology, contributing to the
development of AI-driven voice applications and services.

• The system is capable of real-time age estimation, making it suitable for instant age verification or
personalized content recommendations in real-world scenarios.
 SYSTEM CONFIGURATION

H/W SYSTEM CONFIGURATION:

• PROCESSOR - I3, I5, I7, AMD Processor

• RAM - Above 8 Gb

• HARD DISK - Above 500 GB

S/W SYSTEM CONFIGURATION:

• OPERATING SYSTEM - Windows 8/10

• FRONT END - HTML ,CSS

• NOTEBOOK - Anaconda(Jupiter)

• IDLE - Python(3.10)
 CONCLUSION
• The project has successfully demonstrated the development of a robust and accurate age estimation system
based on human voice using deep neural networks. Through extensive data collection, advanced signal
processing, and the utilization of state-of-the-art neural network architectures, the system can reliably
estimate a person's age from their voice characteristics.

• The achieved high accuracy in age estimation opens up numerous practical applications across healthcare,
entertainment, security, and personalized content delivery. It enhances user experiences by tailoring services
and recommendations based on age, making it a valuable asset for businesses and organizations.

• Ethical considerations have been at the forefront of the project, with a focus on data privacy, consent, and
fairness. The system handles voice data responsibly and complies with legal and regulatory requirements,
ensuring user trust and confidence.In conclusion, the proposed age estimation system represents a significant
technological achievement with practical applications and ethical considerations at its core.

• It stands as a testament to the potential of deep learning and signal processing in voice-related domains while
promoting responsible and ethical AI development.