Extending CLIP for Category-to-image Retrieval in E-commerce
This addresses a specific problem in e-commerce retrieval for improving user sessions, but it is incremental as it builds on existing CLIP technology.
The paper tackles the mismatch between textual and visual category representations in e-commerce by introducing a category-to-image retrieval task and proposing CLIP-ITA, a model that leverages textual, visual, and attribute modalities, which significantly outperforms models using fewer modalities.
E-commerce provides rich multimodal data that is barely leveraged in practice. One aspect of this data is a category tree that is being used in search and recommendation. However, in practice, during a user's session there is often a mismatch between a textual and a visual representation of a given category. Motivated by the problem, we introduce the task of category-to-image retrieval in e-commerce and propose a model for the task, CLIP-ITA. The model leverages information from multiple modalities (textual, visual, and attribute modality) to create product representations. We explore how adding information from multiple modalities (textual, visual, and attribute modality) impacts the model's performance. In particular, we observe that CLIP-ITA significantly outperforms a comparable model that leverages only the visual modality and a comparable model that leverages the visual and attribute modality.