Abstract
Zero-shot learning in Language & Vision is the task of correctly labelling (or naming) objects of novel categories. Another strand of work in L&V aims at pragmatically informative rather than ``correct'' object descriptions, e.g. in reference games. We combine these lines of research and model zero-shot reference games, where a speaker needs to successfully refer to a novel object in an image. Inspired by models of "rational speech acts", we extend a neural generator to become a pragmatic speaker reasoning about uncertain object categories. As a result of this reasoning, the generator produces fewer nouns and names of distractor categories as compared to a literal speaker. We show that this conversational strategy for dealing with novel objects often improves communicative success, in terms of resolution accuracy of an automatic listener.
Abstract (translated)
语言与视觉中的零镜头学习是正确标记(或命名)新类别对象的任务。L&V中的另一部分工作旨在提供实用信息,而不是“正确”的对象描述,例如在参考游戏中。我们结合这些研究线和模型零镜头参考游戏,其中一个演讲者需要成功地参考一个新的对象在一个图像。在“理性言语行为”模型的启发下,我们扩展了一个神经发生器,使之成为关于不确定对象类别的语用说话人推理。这种推理的结果是,与文字演讲者相比,生成器生成的名词和干扰器类别的名称更少。我们表明,这种处理新对象的会话策略通常可以提高自动侦听器的解析精度,从而提高通信成功率。
URL
https://arxiv.org/abs/1906.05518