Productivity Analysis of Employees in Corporate Sector Using Gesture Recognition

In today's dynamic work environments, understanding and optimizing employee productivity are essential for organizational success. This paper introduces an approach for productivity analysis using gesture recognition techniques. By harnessing the capabilities of MediaPipe for body joint detection and LSTM (Long Short-Term Memory) models for action classification, the proposed system enables real-time detection of employee actions. These actions are categorized into four key behaviors: nodding, showing interest, yawning, and resting. Additionally, a mathematical model is introduced to compute overall productivity based on these categorized actions. Moreover, the distribution of actions over time is visualized using graphs, providing valuable insights into employee behavior patterns and facilitating informed decision-making for productivity optimization.