Manufacturing thin-walled structures is challenging due to their high compliance. In this work, the milled workpieces exhibited obvious chatter marks on machined surfaces, which significantly affect surface quality. Traditional chatter detection often relies on complex sensor systems. Although numerical simulations can be used to predict chatter and generate stability lobe diagrams, they require parameters that are difficult to measure. This paper proposes a direct approach based on deep learning and computer vision using YOLOv8 to detect the chatter marks. The performance of the YOLOv8 model was evaluated using several standard fixture components as detection objects. Despite minimal training data, the model demonstrated high reliability. Additionally, this work proposes a deep learning-based method to create stability lobe diagrams.