Landmark-RxR: Solving Vision-and-Language Navigation with Fine-Grained Alignment Supervision | |
Keji He1,2,3; Yan Huang1,2; Qi Wu3; Jianhua Yang5; Dong An1,4; Shuanglin Sima1,2; Liang Wang1,2,6,7 | |
2021-12 | |
会议名称 | Neural Information Processing Systems |
会议日期 | 2021-12-7至2021-12-10 |
会议地点 | 线上 |
摘要 | In Vision-and-Language Navigation (VLN) task, an agent is asked to navigate inside 3D indoor environments following given instructions. Cross-modal alignment is one of the most critical challenges in VLN because the predicted trajectory needs to match the given instruction accurately. In this paper, we address the cross-modal alignment challenge from the perspective of fine-grain. Firstly, to alleviate weak cross-modal alignment supervision from coarse-grained data, we introduce a human-annotated fine-grained VLN dataset, namely Landmark-RxR. Secondly, to further enhance local cross-modal alignment under fine-grained supervision, we investigate the focal-oriented rewards with soft and hard forms, by focusing on the critical points sampled from fine-grained Landmark-RxR. Moreover, to fully evaluate the navigation process, we also propose a re-initialization mechanism that makes metrics insensitive to difficult points, which can cause the agent to deviate from the correct trajectories. Experimental results show that our agent has superior navigation performance on Landmark-RxR, en-RxR and R2R. Our dataset and code are available at https://github.com/hekj/Landmark-RxR. |
收录类别 | EI |
七大方向——子方向分类 | 机器人感知与决策 |
国重实验室规划方向分类 | 多模态协同认知 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/57626 |
专题 | 模式识别实验室 |
作者单位 | 1.Center for Research on Intelligent Perception and Computing National Laboratory of Pattern Recognition Institute of Automation, Chinese Academy of Sciences 2.School of Artificial Intelligence, University of Chinese Academy of Sciences 3.School of Computer Science, University of Adelaide 4.School of Future Technology, University of Chinese Academy of Sciences 5.School of Artificial Intelligence, Beijing University of Posts and Telecommunications 6.Center for Excellence in Brain Science and Intelligence Technology 7.Chinese Academy of Sciences, Artificial Intelligence Research |
第一作者单位 | 中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Keji He,Yan Huang,Qi Wu,et al. Landmark-RxR: Solving Vision-and-Language Navigation with Fine-Grained Alignment Supervision[C],2021. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
NeurIPS-2021-landmar(871KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
个性服务 |
推荐该条目 |
保存到收藏夹 |
查看访问统计 |
导出为Endnote文件 |
谷歌学术 |
谷歌学术中相似的文章 |
[Keji He]的文章 |
[Yan Huang]的文章 |
[Qi Wu]的文章 |
百度学术 |
百度学术中相似的文章 |
[Keji He]的文章 |
[Yan Huang]的文章 |
[Qi Wu]的文章 |
必应学术 |
必应学术中相似的文章 |
[Keji He]的文章 |
[Yan Huang]的文章 |
[Qi Wu]的文章 |
相关权益政策 |
暂无数据 |
收藏/分享 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论