arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1912.01119cs.CVcs.CL
收藏

多个正确输出的深度贝叶斯主动学习

Deep Bayesian Active Learning for Multiple Correct Outputs

Stanford University(斯坦福大学)

Khaled Jedoui, Ranjay Krishna, Michael Bernstein, Li Fei-Fei

摘要:

典型的主动学习策略是为任务设计的,例如分类,假设输出空间是互斥的。这些任务总是有一个正确答案的假设导致了许多基于不确定性的测量,如熵和最小置信度,它们对模型的输出进行操作。不幸的是,许多现实世界的视觉任务,如视觉问答和图像描述,有多个正确的答案,导致这些测量高估了不确定性,有时比随机抽样基线表现更差。在本文中,我们提出了一个新的范式,估计模型的内部隐藏空间,而不是模型的输出空间的不确定性。我们专门研究这个问题的视觉问答生成(VQA),其目的是不分类的正确答案,但产生一个自然语言的答案,给定的图像和问题的表现。我们的方法克服了语言的释义性质。它需要一个语义空间来构建模型的输出概念,并支持使用基于丢弃的贝叶斯不确定性等技术。我们构建了一个视觉语义空间,将任何现有VQA模型的释义紧密地嵌入在一起。我们在两个数据集上展示了VQA任务的最新主动学习结果,Visual Genome的成本效益是VQA 2.0的成本效益的5倍。

英文摘要:

Typical active learning strategies are designed for tasks, such as classification, with the assumption that the output space is mutually exclusive. The assumption that these tasks always have exactly one correct answer has resulted in the creation of numerous uncertainty-based measurements, such as entropy and least confidence, which operate over a model's outputs. Unfortunately, many real-world vision tasks, like visual question answering and image captioning, have multiple correct answers, causing these measurements to overestimate uncertainty and sometimes perform worse than a random sampling baseline. In this paper, we propose a new paradigm that estimates uncertainty in the model's internal hidden space instead of the model's output space. We specifically study a manifestation of this problem for visual question answer generation (VQA), where the aim is not to classify the correct answer but to produce a natural language answer, given an image and a question. Our method overcomes the paraphrastic nature of language. It requires a semantic space that structures the model's output concepts and that enables the usage of techniques like dropout-based Bayesian uncertainty. We build a visual-semantic space that embeds paraphrases close together for any existing VQA model. We empirically show state-of-art active learning results on the task of VQA on two datasets, being 5 times more cost-efficient on Visual Genome and 3 times more cost-efficient on VQA 2.0.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑