Visual impairment significantly affects individuals’ independence and mobility, making daily navigation challenging. This paper presents an assistive mobile application leveraging Artificial Intelligence (AI) for real-time object detection and recognition, integrated with voice feedback to enhance accessibility. The application employs the YOLO (You Only Look Once) algorithm, trained on a diverse dataset to ensure accurate detection in various environments. A text-to-speech (TTS) system provides real-time audio descriptions, allowing users to receive essential information about their surroundings. To optimize performance, the system is deployed on mobile devices using TensorFlow Lite, ensuring efficient on-device inference with minimal latency. Extensive testing evaluates accuracy, response time, and usability, demonstrating high object recognition performance across different scenarios. Results show that the system operates effectively in both indoor and outdoor environments, adapting to varying lighting conditions and object types. Additionally, the lightweight implementation ensures that the application runs smoothly on consumer-grade smartphones, making it an accessible and cost-effective solution. The proposed approach contributes to advancing AI-driven assistive technologies, offering a scalable, user-friendly, and practical tool that empowers visually impaired individuals to navigate their surroundings with greater confidence and autonomy. This study highlights the transformative potential of AI in enhancing accessibility and inclusion, paving the way for future advancements in smart assistive systems.