The study of human regular tasks have become more prevalent and accessible as a result of the widespread use of different sensors integrated into mobile devices. This issue now exists in different vast real-world area applications some examples are “healthcare monitoring, fitness tracking, and user-adaptive systems”. Where a generic model capable of recognizing an arbitrary user’s activity in real-time is necessary. This paper presents a user-independent deep learning technique for digital human activity categorization. The proposed research study advocates the utilization of CNN in combination with basic statistical characteristics to maintain the information regarding the global shape of different time series for performing a local feature extraction. This research study also focuses on how the duration of a time series affects recognition accuracy ans restricting it to a single second to allow time series analysis for performing activity classification.