Automated Red Circle Badge Detection and Clearing for Enhanced Mobile User Experience

With the rapid proliferation of smartphones, red circle badges have become a common visual cue used by mobile applications to notify users of new notifications. While effective in driving user engagement, accumulating these red circles can lead to information overload and user stress. This paper presents a novel solution to automatically clear red circle badges to reduce user stress and improve the mobile user experience. The proposed approach includes two methods: YOLOv7-based model training for detecting red circle badges and a depth-first traversal algorithm for clearing them. Using a dataset of 126 screenshots containing 390 red circle badges, experimental results demonstrate the model achieves high accuracy in detection, with a recall of 98% and a precision of 92%, confirming its robustness. Additionally, experiments conducted on 10 apps show that the red circle badge clearing algorithm performs excellently in terms of efficiency, with an average processing time of 5.9 seconds per badge.