Several mobile applications have been developed to facilitate denomination detection for blind users. However, none of the existing applications allow for detecting multiple notes in a single frame and relaying the total denomination, nor is there a dataset available for the new Indian currency notes, annotated for object detection training. We describe the development of a detection application that aims to improve on the previously existing solutions by enabling multi-note detection, continuous audio feedback, automatic torch usage, and minimal user-application interaction. YOLOv4 allowed the training of a lightweight and fast object detection model with high accuracy on a custom-created dataset post-demonetization of Indian currencies that is deployed on a mobile device.