Fully Data-Driven Pseudo Label Estimation for Pointly-Supervised Panoptic Segmentation | |
Li, Jing1,2,4![]() ![]() ![]() | |
2024 | |
会议名称 | Thirty-Eighth AAAI Conference on Artificial Intelligence (AAAI) |
会议日期 | 2024.2.22-2024.2.25 |
会议地点 | Vancouver, Canada |
摘要 | The core of pointly-supervised panoptic segmentation is estimating accurate dense pseudo labels from sparse point labels to train the panoptic head. Previous works generate pseudo labels mainly based on hand-crafted rules, such as connecting multiple points into polygon masks, or assigning the label information of labeled pixels to unlabeled pixels based on the artificially defined traversing distance. The accuracy of pseudo labels is limited by the quality of the hand-crafted rules (polygon masks are rough at object contour regions, and the traversing distance error will result in wrong pseudo labels). To overcome the limitation of hand-crafted rules, we |
关键词 | Panoptic Segmentation Pointly-Supervised Pseudo Label Estimation Data-Driven |
学科门类 | 工学::计算机科学与技术(可授工学、理学学位) |
收录类别 | EI |
语种 | 英语 |
是否为代表性论文 | 是 |
七大方向——子方向分类 | 图像视频处理与分析 |
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 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/58783 |
专题 | 模式识别实验室 |
通讯作者 | Xiao, Jun |
作者单位 | 1.University of Chinese Academy of Sciences (UCAS) 2.Institute of Automation, Chinese Academy of Sciences (CASIA) 3.Centre for Artificial Intelligence and Robotics, HKISI CAS 4.State Key Laboratory of Multimodal Artificial Intelligence Systems (MAIS) 5.Tencent Maps, Tencent |
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 | Li, Jing,Fan, Junsong,Yang, Yuran,et al. Fully Data-Driven Pseudo Label Estimation for Pointly-Supervised Panoptic Segmentation[C],2024. |
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 | ||
aaai.pdf(2720KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
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