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 2103.11537cs.LGcs.CV
收藏

如何设计样本和计算效率高的VQA模型

How to Design Sample and Computationally Efficient VQA Models

Georgia Tech(佐治亚理工学院);Shanghai Jiao Tong University(上海交通大学);Mohamed bin Zayed University of AI(穆罕默德·本·扎耶德人工智能大学)

Karan Samel, Zelin Zhao, Binghong Chen, Kuan Wang, Robin Luo, Le Song

摘要:

在多模态推理任务中,例如视觉问答(VQA),已经有许多建模和训练范例进行了测试。以前的模型为视觉和语言任务提出了不同的方法,但是哪些方法在样本和计算效率方面表现最好?基于我们的实验,我们发现,表示的文本作为概率程序和图像的对象级场景图最好满足这些desiderata。我们扩展现有的模型,利用这些软程序和场景图训练的问题答案对端到端的方式。实证结果表明,这种可区分的端到端程序执行器能够保持最先进的准确性,同时具有样本和计算效率。

英文摘要:

In multi-modal reasoning tasks, such as visual question answering (VQA), there have been many modeling and training paradigms tested. Previous models propose different methods for the vision and language tasks, but which ones perform the best while being sample and computationally efficient? Based on our experiments, we find that representing the text as probabilistic programs and images as object-level scene graphs best satisfy these desiderata. We extend existing models to leverage these soft programs and scene graphs to train on question answer pairs in an end-to-end manner. Empirical results demonstrate that this differentiable end-to-end program executor is able to maintain state-of-the-art accuracy while being sample and computationally efficient.

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。

↑