Proper chewing is essential for digestion, nutrient absorption, and appetite regulation. However, modern fast-paced lifestyles often lead to rushed eating and insufficient mastication. While previous research has explored methods for sensing eating behavior, few interventions actively promote mindful chewing. This study introduces Eat2pic-Mobile, an interactive system designed to encourage thorough chewing through Chew-Draw, a visual interaction that translates chewing activity into a digital painting. Unlike previous approaches that rely on specialized sensors, Eat2pic-Mobile uses commercially available smartphones and wireless earphones to detect chewing sounds and classify food types via machine learning. A user study with 30 participants suggested that Eat2pic-Mobile may enhance awareness of chewing and encourage more balanced meal choices. The art-based visualization was found to be significantly more engaging than a conventional graph-based representation and was also perceived as more accurate, despite both using the same underlying machine learning model. Qualitative feedback indicated that the artistic representation encouraged users to chew more thoroughly and pay greater attention to dietary balance, highlighting the potential of aesthetic feedback in behavior change interventions. These findings indicate that interactive visual feedback may support mindful eating and could offer possibilities for scalable, technology-driven dietary interventions.