Accessible Intelligence: Deep Learning-Driven Voice Assistant Interface for Diverse Disabilities
摘要
The voice assistant presented in this paper is intended to increase accessibility and inclusion for people with disabilities, such as those who have cognitive, hearing, or vision problems. To assist users with their everyday tasks, the system incorporates technologies including computer vision, facial identification, emotion analysis, sign language detection, and speech recognition. By identifying Indian Sign Language (ISL)’s action-based gestures as well as its alphabet, it provides individualized help, removing barriers to communication for sign language users and improving accessibility. In addition, the assistant lets users manage their laptops with voice commands or sign language motions, make playlists of music depending on moods, summarize news and information related to their interests, help with everyday activities via dialogue, and provide navigation support. By adapting to the user’s abilities, preferences, and emotional states, this system promotes independence and ease of use for individuals with disabilities who often rely on continuous support. This research advances the development of intelligent systems that foster inclusion and empower people with disabilities.