Catalog Phrase Grounding (CPG): Grounding of Product Textual Attributes in Product Images for e-commerce Vision-Language Applications
This addresses the need for better vision-language integration in e-commerce catalog systems, though it is incremental as it builds on existing methods like MDETR and logo detection models.
The paper tackles the problem of associating product textual attributes with corresponding regions in product images for e-commerce applications, using a self-supervised model trained on 2.3 million image-text pairs, which improves recall by 5% on average in product-brand matching at fixed precision.
We present Catalog Phrase Grounding (CPG), a model that can associate product textual data (title, brands) into corresponding regions of product images (isolated product region, brand logo region) for e-commerce vision-language applications. We use a state-of-the-art modulated multimodal transformer encoder-decoder architecture unifying object detection and phrase-grounding. We train the model in self-supervised fashion with 2.3 million image-text pairs synthesized from an e-commerce site. The self-supervision data is annotated with high-confidence pseudo-labels generated with a combination of teacher models: a pre-trained general domain phrase grounding model (e.g. MDETR) and a specialized logo detection model. This allows CPG, as a student model, to benefit from transfer knowledge from these base models combining general-domain knowledge and specialized knowledge. Beyond immediate catalog phrase grounding tasks, we can benefit from CPG representations by incorporating them as ML features into downstream catalog applications that require deep semantic understanding of products. Our experiments on product-brand matching, a challenging e-commerce application, show that incorporating CPG representations into the existing production ensemble system leads to on average 5% recall improvement across all countries globally (with the largest lift of 11% in a single country) at fixed 95% precision, outperforming other alternatives including a logo detection teacher model and ResNet50.