Maesh Conducts Visitor AI Analysis PoC at LG Uplus Multi-Cultural Space

Video processing AI startup Mayi is partnering with LG Uplus to open a new horizon for offline space operation. The two companies have signed a Proof of Concept (PoC) agreement to introduce Mayi's innovative visitor analysis AI solution, 'mAsh,' into LG Uplus' complex cultural space 'Everyday_Unusual byU+' by May. 'Everyday_Unusual byU+', an MZ generation hot spot at Gangnam Station, sees explosive foot traffic. LG Uplus aims to quantitatively measure and optimize store performance and customer experience through mAsh. By utilizing staff uniforms and movement path information, visitors will be precisely distinguished, and diverse data, including visitor count, gender, age group, and even consultation conversion rates, will be secured to maximize the store's potential. mAsh's 'Smart Tracking Plan' meticulously tracks not only entry/exit data such as foot traffic and visit conversion rates but also visitors' main movement paths and dwell times, providing key insights for space operation through an intuitive dashboard. By analyzing CCTV footage installed in offline spaces, mAsh presents Key Performance Indicators (KPIs) tailored to each store's operational goals and essential data for marketing strategy development, supporting efficient space operation. At the heart of all this analysis is Mayi's unique 're-identification technology.' This technology accurately recognizes visitors captured in various zones as the same individual and perfectly distinguishes them from staff information. With an astonishing 92% accuracy that surpasses the latest academic models, it fundamentally prevents errors in duplicate data aggregation, establishing a more reliable foundation than any other data analysis. Kim Chan-gyu, CEO of Mayi, expressed his aspirations for the future, stating, "Through collaboration with LG Uplus, we will lead data-driven store optimization and customer experience innovation." Meanwhile, Mayi's re-identification technology research was selected as a 'Spotlight' recipient, awarded to the top 5% at ICLR 2025, the world's most prestigious AI academic conference, once again proving its unrivaled technological prowess on the global stage.