Gender Classification and Age Detection Syste...

Gender Classification and Age Detection System Using CNN on Real-Time Video Crowd

Author Name : Miss. Dimpal A. Kapgate, Prof. Abhimanyu Dutonde, Prof. Pragati Patil

ABSTRACT The essential need for automatic facial image-based age and gender classification has grown since the emergence of social media, the banking sector, medical sector, industry, market area also, and the Verification Process of government identification cards. Therefore, age and gender classification are an important step in many applications, including interest group targeting, aging analysis, face verification, and ad targeting. Based on the existing works, there exist various limitations and research gaps in Gender Classification and Age Detection systems on Real-Time Video in Crowd areas. In most cases, the age range is estimated instead of the accurate age of a person. In this research, we discuss the method that improves the accuracy of the existing methodology. There are also works on addressing the age estimation of multiple people in front of a camera also detecting the people's age and gender at a time on real-time video. Deep learning techniques are useful for a range of tasks, including gender and age prediction, object recognition, feature extraction, and classification. Using this technology, we were able to recognize the age and gender of each image in real-time video capture using CCTV of numerous photographs. This study will describe a Convolutional Neural Network (CNNs) that uses deep learning, as well as applicable techniques and algorithms, and how everything works together to classify gender and detect age. Using this technology, we were able to recognize the age and gender of each image in real-time video of numerous photographs at once. Variations may not occur due to environmental lighting conditions, head pose, hat, specs, or goggles in the performance of the gender classification algorithm in terms of classification rate. Due to the minor difference between the faces of males and females is difficult to identify in real-time video numerous photographs at a time. Using the Geometric-based and Appearance-based methods it May be possible to identify.

Keywords: Convolutional Neural Networks (CNNs), deep learning, gender classification, age detection.