Learning Visual Grounding from Generative Vision and Language Model
This work addresses the data bottleneck in visual grounding for computer vision researchers by creating a large-scale dataset using model-generated queries, which is incremental as it builds on existing generative models and object detection data.
The authors tackled the problem of scaling up text annotation for visual grounding by leveraging generative vision-language models to generate object-level descriptions from object detection datasets, resulting in a dataset of 500K images and 16M referring expressions that significantly outperforms state-of-the-art methods on RefCOCO benchmarks without human-annotated data.
Visual grounding tasks aim to localize image regions based on natural language references. In this work, we explore whether generative VLMs predominantly trained on image-text data could be leveraged to scale up the text annotation of visual grounding data. We find that grounding knowledge already exists in generative VLM and can be elicited by proper prompting. We thus prompt a VLM to generate object-level descriptions by feeding it object regions from existing object detection datasets. We further propose attribute modeling to explicitly capture the important object attributes, and spatial relation modeling to capture inter-object relationship, both of which are common linguistic pattern in referring expression. Our constructed dataset (500K images, 1M objects, 16M referring expressions) is one of the largest grounding datasets to date, and the first grounding dataset with purely model-generated queries and human-annotated objects. To verify the quality of this data, we conduct zero-shot transfer experiments to the popular RefCOCO benchmarks for both referring expression comprehension (REC) and segmentation (RES) tasks. On both tasks, our model significantly outperform the state-of-the-art approaches without using human annotated visual grounding data. Our results demonstrate the promise of generative VLM to scale up visual grounding in the real world. Code and models will be released.