South KoreaRetail and e-commerce2020 – ongoingSamsung C&T
AI outfit recommendations on SSF Shop
Samsung C&T deployed AI outfit recommendations that POSTECH said replaced more than 30 hours of processing with real-time results
Problem
Online shoppers lacked the styling assistance available in stores. Fashion models trained on overseas data also struggled to reflect Korean styles.
Approach
Samsung C&T and POSTECH developed a recommendation service over a year before its 2021 announcement. They trained it on outfits assembled with fashion experts and used Samsung C&T's visual AI to extract clothing attributes. The service went live on SSF Shop, recommending coordinated items around a customer's selected garment.
Results
Takeaway
Brand experts supplied the outfit combinations used for training. POSTECH reported faster recommendation processing, without disclosing a sales effect.
Categories
- Size
- Large enterprise
- Industry
- Retail and e-commerce
- Business function
- Marketing and sales
- Innovation type
- Product and service innovation
- AI technology
- Recommendation and personalizationComputer vision
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
POSTECH described its jointly developed fashion recommendation service already operating on SSF Shop. It reported that processing previously taking more than 30 hours could be performed in real time.
Full details are in the original sources.