Consistency Regularization for Complementary Clothing Recommendations

Abstract

This paper reports on the development of a Consistency Regularized model for Bayesian Personalized Ranking (CR-BPR), addressing to the drawbacks in existing complementary clothing recommendation methods, namely limited consistency and biased learning caused by diverse feature scale of multi-modal data. Compared to other product types, fashion preferences are inherently subjective and more personal, and fashion are often presented, not by individual clothing product, but with other complementary product(s) in a well coordinated fashion outfit. Current complementary-product recommendation studies primarily focus on user preference and product matching, this study further emphasizes the consistency observed in user-product interactions as well as product-product interactions, in the specific context of clothing matching. Most traditional approaches often underplayed the impact of existing wardrobe items on future matching choices, resulting in less effective preference prediction models. Moreover, many multi-modal information based models overlook the limitations arising from various feature scales being involved. To address these gaps, the CR-BPR model integrates collaborative filtering techniques to incorporate both user preference and product matching modeling, with a unique focus on consistency regularization for each aspect. Additionally, the incorporation of a feature scaling process further addresses the imbalances caused by different feature scales, ensuring that the model can effectively handle multi-modal data without being skewed by any particular type of feature. The effectiveness of the CR-BPR model was validated through detailed analysis involving two benchmark datasets. The results confirmed that the proposed approach significantly outperforms existing models.

Figures (7)

• Figure 1: Key concepts of personalized clothing matching
• Figure 2: The CR-BPR scheme has four main components. User preferences are predicted by user-product interactions. Product Matching Prediction module uses visual and textual product-product interaction principles. The User Preference Consistency regularization module captures the average similarity between the target matching garment and user historical choices on both visual and textual modalities. The Product Matching Consistency regularization module determines the mean similarity between the target matching garment and the historical choices matched for the given clothing.
• Figure 3: The performance comparison between the GP-BPR baseline and GP-BPR with Feature Scaling using the Polyvore-519 and IQON3000 datasets when (a) specify top and recommend bottom, and (b) specify bottom and recommend top. $\mu$ is the weight of Product Matching modeling.
• Figure 4: Illustration of the clothing matching recommendation results provided by three methods. The ground-truth are highlighted by red frame.
• Figure 5: Comparison of the clothing matching recommendation results provided by different methods.