This study explores the impact of implementing a deeplearning-based approach in cultural arts education, specifically in batik-making lessons at the senior high school level. The problem addressed in this research is the lack of student engagement and interest in conventional teaching methods, which often results in lower comprehension and artwork quality. The aim of this study is to compare the understanding, artwork quality, and enthusiasm between students in the experimental group, who received practice-first instruction followed by theoretical content, and those in the control group, who followed a traditional learning approach. A quantitative method was applied by utilizing pre-test and post-test questionnaires, interviews, and direct observation in the classroom. Knowledge was measured through multiple-choice questions, feelings were gauged via interviews, and student actions were observed during the learning process. Results show a significant improvement in the experimental class, with higher post-test scores and more positive attitudes toward batik-making. In contrast, the control class exhibited lower enthusiasm and engagement, which subsequently affected their final grades. The findings suggest that deeplearning-based methods can provide a more impactful and memorable learning experience, promoting a deeper understanding and appreciation of cultural heritage. Further research is recommended to explore the application of this method in other cultural arts disciplines and assess its long-term effectiveness in developing students' cognitive and creative skills.