Continual Stereo Matching of Continuous Driving Scenes with Growing Architecture
Zhang, Chenghao1,2; Tian, Kun1,2; Fan, Bin3; Meng, Gaofeng1,2,4; Zhang, Zhaoxiang1,4; Pan, Chunhong1
2022-06
会议名称IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
会议日期2022.06.19
会议地点美国路易斯安那州新奥尔良
摘要
The deep stereo models have achieved state-of-the-art performance on driving scenes, but they suffer from severe performance degradation when tested on unseen scenes. Although recent work has narrowed this performance gap through continuous online adaptation, this setup requires continuous gradient updates at inference and can hardly deal with rapidly changing scenes. To address these challenges, we propose to perform continual stereo matching where a model is tasked to 1) continually learn new scenes, 2) overcome forgetting previously learned scenes, and 3) continuously predict disparities at deployment. We achieve this goal by introducing a Reusable Architecture Growth (RAG) framework. RAG leverages task-specific neural unit search and architecture growth for continual learning of new scenes. During growth, it can maintain high reusability by reusing previous neural units while achieving good performance. A module named Scene Router is further introduced to adaptively select the scene-specific architecture path at inference. Experimental results demonstrate that our method achieves compelling performance in various types of challenging driving scenes.
语种英语
七大方向——子方向分类三维视觉
国重实验室规划方向分类视觉信息处理
是否有论文关联数据集需要存交
文献类型会议论文
条目标识符http://ir.ia.ac.cn/handle/173211/51491
专题模式识别国家重点实验室_先进时空数据分析与学习
通讯作者Meng, Gaofeng
作者单位1.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 Automation and Electrical Engineering, University of Science and Technology Beijing
4.CAS Centre for Artificial Intelligence and Robotics, HK Institute of Science and Innovation
第一作者单位模式识别国家重点实验室
通讯作者单位模式识别国家重点实验室
推荐引用方式
GB/T 7714
Zhang, Chenghao,Tian, Kun,Fan, Bin,et al. Continual Stereo Matching of Continuous Driving Scenes with Growing Architecture[C],2022.
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
论文序号1_CVPR.pdf(2647KB)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
[Zhang, Chenghao]的文章
[Tian, Kun]的文章
[Fan, Bin]的文章
百度学术
百度学术中相似的文章
[Zhang, Chenghao]的文章
[Tian, Kun]的文章
[Fan, Bin]的文章
必应学术
必应学术中相似的文章
[Zhang, Chenghao]的文章
[Tian, Kun]的文章
[Fan, Bin]的文章
相关权益政策
暂无数据
收藏/分享
文件名: 论文序号1_CVPR.pdf
格式: Adobe PDF
所有评论 (0)
暂无评论
 

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。