Publication

Abstract : Automated attendance monitoring is essential across various fields, offering efficient and precise alternatives to traditional manual methods. This paper presents an innovative approach to attendance tracking by utilizing a Raspberry Pi, microphone, audio input, and machine learning techniques. The system integrates LSTM-based ambient sensitivity and RSSI-based proximity analysis to achieve heightened accuracy in attendance monitoring. The Raspberry Pi acts as the central component, capturing ambient audio with a microphone and collecting RSSI data via a $\mathrm{Wi}-\mathrm{Fi}$ module. MFCC is employed for audio preprocessing to extract relevant features for training an LSTM model, capable of classifying diverse audio inputs. Simultaneously, RSSI data undergoes preprocessing and analysis to estimate device proximity accurately using machine learning algorithms. This paper presents a particular combination of current methods (RSSI, LSTM, and Raspberry Pi) for dual-audio attendance monitoring. With two distinct audio inputs for the teacher and pupils, it may be possible to analyse the presence of the teacher differently and to discern ambient noise based on location. By comparing audio inputs quantitatively, this method gives the attendance monitoring system an extra degree of accuracy. The proposed system’s student module records attendance by analyzing ambient sound and RSSI data processed by LSTM with MFCC. This comprehensive approach reduces manual efforts, enhances accuracy, and delivers real-time attendance information to the teacher module, facilitating efficient attendance management in educational environments.