Voice-Activated Smart Homes: Enhancing Accessibilty for Non-Literate Individuals

The expeditious integration of technology into our daily life has developed from simple switches to complicated AI systems that are significantly influencing daily routines. Setting apart their benefits, the current voice-based AI system lacks precision and adaptability, specifically in accurately interpreting user commands across various accents and dialects. This paper introduces an enhanced voice-activated smart home control system, focusing on non-literate users by addressing limitations in adaptive voice recognition, privacy preserving data processing and efficient ML models that significantly improvises the system performance from pre-existing ones. Key contributions include an adaptive voice recognition model that adjusts to user-specific accents and dialects for various slangs, an edge-based approach to data-processing to strengthen privacy. Additionally, this system inculcates context-aware device management and tone recognition in the voice commands of the users, to adaptively respond to their intent and routine, enhancing their daily usability, security and accessibility. The proposed system offers a dependable user centric advancement in smart home automation that fills the void between accessibility for diverse and non-literate user groups and technology itself.