Technological advancements in Artificial Intelligence and Human–Computer Interaction has revolutionized assistive technology for the upliftment of the Blind and Low Vision Individuals (BLVIs). They have provided them with safe and supportive mobility, navigation, and emotional stability. While advancements in AI-enabled assistive technology have paved the way for new research paradigms, the ability of HAR to provide environmental awareness for BLVIs through assistive technology is unexplored. By incorporating sensor-based data in glasses and using pervasive computing we propose an integrated system of AI-enabled smart glasses with HAR and proximity sensors for the BLVIs for better navigation in indoor environments. This paper proposes a deep learning-based HAR algorithm on UCA-EHAR dataset with an accuracy of 96.3% to predict the activities specific to walking. It also includes distance calculation of visible objects through proximity sensors and, finally converting the results to audio for BLVIs. This integrated system will provide real-time auditory feedback about the complex indoor environment and foster greater independence.