CapWAP: Captioning with a Purpose
This addresses the limitation of one-size-fits-all image captioning for diverse user populations, offering a more targeted approach, though it is incremental as it builds on existing captioning and VQA methods.
The paper tackles the problem of generic image captioning by proposing a new task, Captioning with a Purpose (CapWAP), which tailors captions to specific user information needs using question-answer pairs, and demonstrates that their system outperforms generic captioning methods in fulfilling these needs as measured by QA performance on unseen images.
The traditional image captioning task uses generic reference captions to provide textual information about images. Different user populations, however, will care about different visual aspects of images. In this paper, we propose a new task, Captioning with a Purpose (CapWAP). Our goal is to develop systems that can be tailored to be useful for the information needs of an intended population, rather than merely provide generic information about an image. In this task, we use question-answer (QA) pairs---a natural expression of information need---from users, instead of reference captions, for both training and post-inference evaluation. We show that it is possible to use reinforcement learning to directly optimize for the intended information need, by rewarding outputs that allow a question answering model to provide correct answers to sampled user questions. We convert several visual question answering datasets into CapWAP datasets, and demonstrate that under a variety of scenarios our purposeful captioning system learns to anticipate and fulfill specific information needs better than its generic counterparts, as measured by QA performance on user questions from unseen images, when using the caption alone as context.