Continual Stereo Matching of Continuous Driving Scenes with Growing Architecture
文献类型:会议论文
作者 | Zhang, Chenghao2,4; Tian, Kun2,4; Fan, Bin1; Meng, Gaofeng2,3,4; Zhang, Zhaoxiang2,3; Pan, Chunhong2 |
出版日期 | 2022-06 |
会议日期 | 2022.06.19 |
会议地点 | 美国路易斯安那州新奥尔良 |
英文摘要 | 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 |
源URL | [http://ir.ia.ac.cn/handle/173211/51491] |
专题 | 自动化研究所_模式识别国家重点实验室_遥感图像处理团队 |
通讯作者 | Meng, Gaofeng |
作者单位 | 1.School of Automation and Electrical Engineering, University of Science and Technology Beijing 2.National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences 3.CAS Centre for Artificial Intelligence and Robotics, HK Institute of Science and Innovation 4.School of Artificial Intelligence, University of Chinese Academy of Sciences |
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]. 见:. 美国路易斯安那州新奥尔良. 2022.06.19. |
入库方式: OAI收割
来源:自动化研究所
浏览0
下载0
收藏0
其他版本
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。