Towards Unifying Reference Expression Generation and Comprehension
This work addresses the problem of improving both REG and REC tasks in computer vision and natural language processing, but it is incremental as it builds on existing joint modeling approaches.
The paper tackled the joint modeling of Reference Expression Generation (REG) and Comprehension (REC) by proposing UniRef, a unified model with an Image-Region-Text Fusion layer and pre-training tasks, which achieved state-of-the-art performance on benchmark datasets like RefCOCO, RefCOCO+, and RefCOCOg.
Reference Expression Generation (REG) and Comprehension (REC) are two highly correlated tasks. Modeling REG and REC simultaneously for utilizing the relation between them is a promising way to improve both. However, the problem of distinct inputs, as well as building connections between them in a single model, brings challenges to the design and training of the joint model. To address the problems, we propose a unified model for REG and REC, named UniRef. It unifies these two tasks with the carefully-designed Image-Region-Text Fusion layer (IRTF), which fuses the image, region and text via the image cross-attention and region cross-attention. Additionally, IRTF could generate pseudo input regions for the REC task to enable a uniform way for sharing the identical representation space across the REC and REG. We further propose Vision-conditioned Masked Language Modeling (VMLM) and Text-Conditioned Region Prediction (TRP) to pre-train UniRef model on multi-granular corpora. The VMLM and TRP are directly related to REG and REC, respectively, but could help each other. We conduct extensive experiments on three benchmark datasets, RefCOCO, RefCOCO+ and RefCOCOg. Experimental results show that our model outperforms previous state-of-the-art methods on both REG and REC.