arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1910.03343cs.CVcs.CLcs.LG
收藏

VQA的调制自注意卷积网络

Modulated Self-attention Convolutional Network for VQA

Polytechnic Mons(蒙斯理工学院);Telecom SudParis(巴黎南泰尔电信学院)

Jean-Benoit Delbrouck, Antoine Maiorca, Nathan Hubens, Stéphane Dupont

摘要:

随着用于真实世界视觉推理和组合问答的新数据集的出现,可能需要在训练期间使用视觉特征提取作为端到端过程。这一小小的贡献旨在提出新的想法,以改善传统的卷积网络的视觉处理的视觉问答(VQA)。在本文中,我们提出了调制的语言输入的CNN增强自我注意。我们在这一方向的未来研究显示出令人鼓舞的相对改善。

英文摘要:

As new data-sets for real-world visual reasoning and compositional question answering are emerging, it might be needed to use the visual feature extraction as a end-to-end process during training. This small contribution aims to suggest new ideas to improve the visual processing of traditional convolutional network for visual question answering (VQA). In this paper, we propose to modulate by a linguistic input a CNN augmented with self-attention. We show encouraging relative improvements for future research in this direction.

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。

↑