A Machine Learning Based Image to Object Detection and Audio Description for Blind People

Autonomous driving may enhance road safety, traffic congestion, and the environment. However, several obstacles prevent autonomous vehicle adoption. The work converts visuals into audio for blind people. The visually challenged need this project because it uses machine learning to recognize objects in photographs and provide audio descriptions, bridging the accessibility gap and giving them meaningful access to visual information. This program uses machine learning to translate visual material into aural information to help blind people feel included and independent in a more equal society. The Image Collection, carefully selected, has over 8,000 different photos with item names and audio explanations to help the visually handicapped. The hybrid strategy combines VGG-16, CNN, and NLP. CNN identifies things quickly, VGG-16 classifies well, and NLP synthesizes speech. Multimodal transformation effortlessly synthesizes selected text into audio descriptions, harmonizing modalities for accessibility after training and testing. Visualizing the design using a VGG16 model graph yields 95.80% recognition accuracy. This breakthrough and smart glasses may revolutionize object detection and give the blind control over their lives. It stands out amid a sea of efforts and shows a genuine commitment to help the blind.