A Generative Approach for Wikipedia-Scale Visual Entity Recognition
This addresses the problem of large-scale visual recognition for applications like search and knowledge bases, presenting a novel method rather than an incremental improvement.
The paper tackles web-scale visual entity recognition by mapping query images to one of 6 million Wikipedia entities, introducing a Generative Entity Recognition (GER) framework that auto-regressively decodes entity codes and achieves state-of-the-art performance on the OVEN benchmark.
In this paper, we address web-scale visual entity recognition, specifically the task of mapping a given query image to one of the 6 million existing entities in Wikipedia. One way of approaching a problem of such scale is using dual-encoder models (eg CLIP), where all the entity names and query images are embedded into a unified space, paving the way for an approximate k-NN search. Alternatively, it is also possible to re-purpose a captioning model to directly generate the entity names for a given image. In contrast, we introduce a novel Generative Entity Recognition (GER) framework, which given an input image learns to auto-regressively decode a semantic and discriminative ``code'' identifying the target entity. Our experiments demonstrate the efficacy of this GER paradigm, showcasing state-of-the-art performance on the challenging OVEN benchmark. GER surpasses strong captioning, dual-encoder, visual matching and hierarchical classification baselines, affirming its advantage in tackling the complexities of web-scale recognition.