机械工程学报 ›› 2026, Vol. 62 ›› Issue (17): 257-268.doi: 10.3901/JME.260729
• 特邀专栏:人形机器人与具身智能 • 上一篇
张嘉东, 王巍
收稿日期:2025-12-31
修回日期:2026-06-01
发布日期:2026-11-20
作者简介:张嘉东,男,1996年出生,博士研究生。主要研究方向为语义建图、物体建模。E-mail:zhangjiadong@buaa.edu.cn;王巍(通信作者),男,1972年出生,博士,教授,博士研究生导师。主要研究方向为机器人定位与建图、机器人控制与规划。E-mail:wangweilab@buaa.edu.cn
基金资助:ZHANG Jiadong, WANG Wei
Received:2025-12-31
Revised:2026-06-01
Published:2026-11-20
摘要: 精确且完整的语义地图对于物体搜索至关重要。然而在真实室内环境中,由于物体遮挡、机器人视野受限、空间布局多样化等因素,语义地图预测与物体搜索的性能被严重限制。针对上述问题,提出一种基于关系推理和主动训练的语义地图预测方法。首先,构建栅格级语义空间关系推理模块,通过语义关系图和空间关系图建模物体类别共现规律与空间邻近关系,从而增强未观测区域的语义补全能力。其次,提出基于信息增益的主动训练策略,利用集成模型估计各栅格的认知不确定性,引导机器人主动采集高价值训练样本,以提升模型在未知环境中的泛化能力。最后,设计不确定性驱动的物体搜索机制,通过置信上界评分函数联合考虑目标类别预测概率与不确定性,实现探索与利用的动态平衡。在基于Matterport3D数据集的仿真室内环境中对所提方法进行了验证,实验结果表明在语义建图精度与物体搜索性能方面均优于对比方法。
中图分类号:
张嘉东, 王巍. 基于关系推理和主动训练的室内语义地图预测方法[J]. 机械工程学报, 2026, 62(17): 257-268.
ZHANG Jiadong, WANG Wei. Indoor Semantic Map Prediction Based on Relationship Reasoning and Active Training[J]. Journal of Mechanical Engineering, 2026, 62(17): 257-268.
| [1] 阴贺生,裴硕,徐磊,等. 多机器人视觉同时定位与建图技术研究综述[J]. 机械工程学报,2022,58(11):11-36. YIN Hesheng,PEI Shuo,XU Lei,et al. Review of research on multi-robot visual simultaneous localization and mapping[J]. Journal of Mechanical Engineering,2022,58(11):11-36. [2] WANG J,RÜNZ M,AGAPITO L. DSP-SLAM:Object oriented SLAM with deep shape priors[C]//2021 International Conference on 3D Vision (3DV). IEEE,2021:1362-1371. [3] CHAPLOT D S,GANDHI D P,GUPTA A,et al. Object goal navigation using goal-oriented semantic exploration[J]. Advances in Neural Information Processing Systems,2020,33:4247-4258. [4] NARASIMHAN M,WIJMANS E,CHEN X,et al. Seeing the un-scene:Learning amodal semantic maps for room navigation[C]//European Conference on Computer Vision. Cham:Springer International Publishing,2020:513-529. [5] KATYAL K,POPEK K,PAXTON C,et al. Uncertainty-aware occupancy map prediction using generative networks for robot navigation[C]//2019 International Conference on Robotics and Automation (ICRA). IEEE,2019:5453-5459. [6] SHRESTHA R,TIAN F P,FENG W,et al. Learned map prediction for enhanced mobile robot exploration[C]//2019 International Conference on Robotics and Automation (ICRA). IEEE,2019:1197-1204. [7] GEORGAKIS G,BUCHER B,SCHMECKPEPER K,et al. Learning to map for active semantic goal navigation[C]//International Conference on Learning Representations(ICLR),2022. [8] MENG Y,GUO C,LI A,et al. Context-Aware Graph Inference and Generative Adversarial Imitation Learning for Object-Goal Navigation in Unfamiliar Environment[J]. IEEE Robotics and Automation Letters,2025,10(4):3803-3810. [9] DRUON R,YOSHIYASU Y,KANEZAKI A,et al. Visual object search by learning spatial context[J]. IEEE Robotics and Automation Letters,2020,5(2):1279-1286. [10] DU H,YU X,ZHENG L. Learning object relation graph and tentative policy for visual navigation[C]//European Conference on Computer Vision. Cham:Springer International Publishing,2020:19-34. [11] LYU Y,SHI Y,ZHANG X. Improving target-driven visual navigation with attention on 3D spatial relationships[J]. Neural Processing Letters,2022,54(5):3979-3998. [12] GUAN H,LIN J,LAU R W H. Learning semantic associations for mirror detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2022:5941-5950. [13] JIANG B,ZHANG Z,LIN D,et al. Semi-supervised learning with graph learning-convolutional networks[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2019:11313-11320. [14] SETTLES B. Active learning literature survey[D]. Madison:University of Wisconsin-Madison,2009. [15] GUO C,BAI T,WANG X,et al. ShadowPainter:Active learning enabled robotic painting through visual measurement and reproduction of the artistic creation process[J]. Journal of Intelligent & Robotic Systems,2022,105(3):61. [16] BELUCH W H,GENEWEIN T,NÜRNBERGER A,et al. The power of ensembles for active learning in image classification[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018:9368-9377. [17] WANG K,ZHAN B,ZU C,et al. Semi-supervised medical image segmentation via a tripled-uncertainty guided mean teacher model with contrastive learning[J]. Medical Image Analysis,2022,79:102447. [18] JIN Q,YUAN M,QIAO Q,et al. One-shot active learning for image segmentation via contrastive learning and diversity-based sampling[J]. Knowledge-Based Systems,2022,241:108278. [19] LI D,WANG Z,CHEN Y,et al. A survey on deep active learning:Recent advances and new frontiers[J]. IEEE Transactions on Neural Networks and Learning Systems,2024,36(4):5879-5899. [20] YADAV K,RAMRAKHYA R,RAMAKRISHNAN S K,et al. Habitat-matterport 3D semantics dataset[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2023:4927-4936. [21] 蒋林,聂文康,朱建阳,等. 基于具有墙角信息的语义地图改进AMCL重定位算法[J]. 机械工程学报,2022,58(24):312-323. JIANG Lin,NIE Wenkang,ZHU Jianyang,et al. Improved AMCL relocation algorithm based on semantic map with corner information[J]. Journal of Mechanical Engineering,2022,58(24):312-323. [22] RONNEBERGER O,FISCHER P,BROX T. U-net:Convolutional networks for biomedical image segmentation[C]//International Conference on Medical Image Computing and Computer-assisted Intervention. Cham:Springer International Publishing,2015:234-241. [23] KENDALL A,GAL Y. What uncertainties do we need in bayesian deep learning for computer vision?[C]// Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS’17). Curran Associates,Inc.,2017:5580-5590. [24] NALEWAJSKI R F. Elements of information theory[M]. Heidelberg:Springer Berlin Heidelberg,2011. [25] BERNARDO,JOSE M. Expected information as expected utility[J]. Annals of Statistics,1979,7(3):686-690. [26] LAKSHMINARAYANAN B,PRITZEL A,BLUNDELL C. Simple and scalable predictive uncertainty estimation using deep ensembles[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook,NY,USA:Curran Associates Inc.,2017:6405-6416. [27] GANAIE M A,HU M,MALIK A K,et al. Ensemble deep learning:A review[J]. Engineering Applications of Artificial Intelligence,2022,115:105151. [28] AUER P,CESA-BIANCHI N,FISCHER P. Finite-time analysis of the multiarmed bandit problem[J]. Machine Learning,2002,47(2):235-256. [29] SAVVA M,KADIAN A,MAKSYMETS O,et al. Habitat:A platform for embodied AI research[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. 2019:9339-9347. [30] SUN J,WU J,JI Z,et al. RSMPNet:Relationship guided semantic map prediction[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. 2024:303-312. [31] WIJMANS E,KADIAN A,MORCOS A,et al. DD-PPO:Learning near-perfect pointgoal navigators from 2.5 billion frames[J]. arXiv preprint arXiv:1911.00357,2019. [32] WANG J,TARRIO J,AGAPITO L,et al. SeMLaPS:Real-time semantic mapping with latent prior networks and quasi-planar segmentation[J]. IEEE Robotics and Automation Letters,2023,8(12):7954-7961. [33] TAO Y,GAO H,WEN Y,et al. Glass recognition and map optimization method for mobile robot based on boundary guidance[J]. Chinese Journal of Mechanical Engineering,2023,36(1):74. [34] SHENTU S,GONG Z,LIU X J,et al. Hybrid navigation system based autonomous positioning and path planning for mobile robots[J]. Chinese Journal of Mechanical Engineering,2022,35(1):109. |
| [1] | 陈久朋, 罗缨栾, 伞红军, 李春磊, 张帆. 轮足复合式移动机器人结构、控制与仿真研究综述[J]. 机械工程学报, 2026, 62(15): 291-323. |
| [2] | 马跃, 彭盾, 孙伟华, 李彬, 刘祺, 陈龙, 彭虹贵, 杨玉维. 新型两支链单平台四自由度高速并联机器人动力学建模与动力尺度综合[J]. 机械工程学报, 2026, 62(15): 366-381. |
| [3] | 张昕, 史东豪, 陈芝翰, 徐灵敏, 李秦川. 用于微创手术的混联机器人机构设计与轨迹分层优化[J]. 机械工程学报, 2026, 62(11): 201-214. |
| [4] | 张聘, 姜山, 韩成, 白龙, 布天昭, 吴豪. 面向智能机器人的大面积电子皮肤感知系统[J]. 机械工程学报, 2026, 62(11): 241-248. |
| [5] | 陈夏非, 焦翊洋, 姜雪, 孟昭亮, 张立杰, 刘小平. 基于自适应模型预测控制的驾驶模拟器运动提示算法设计与验证[J]. 机械工程学报, 2026, 62(10): 298-307. |
| [6] | 蒋沁诚, 陶建峰, 汪世杰, 王洋洋, 刘成良. 基于TCN-Transformer的工业机器人自动异常检测[J]. 机械工程学报, 2026, 62(9): 1-13. |
| [7] | 崔健锋, 窦嘉琳, 姜洪洲. 计及负载变化的容积驱动人工肌肉长度预测方法[J]. 机械工程学报, 2026, 62(9): 14-25. |
| [8] | 赵智远, 赵京东, 杨晓航, 梁西昌, 王传英, 万熠. 基于被动式伸缩臂杆的SSRMS构型可重构空间机械臂重构控制策略[J]. 机械工程学报, 2026, 62(9): 26-41. |
| [9] | 樊波, 胡波, 陈子明, 姚建涛. 基于非均匀有理B样条的无人装载机铲装轨迹多目标优化[J]. 机械工程学报, 2026, 62(7): 185-195. |
| [10] | 顾连睿, 郭国强, 江仁政, 毋天歌, 柯帅, 孔志学, 赵欢. 薄壁复杂结构的机器人打磨路径规划和力位控制技术现状与发展[J]. 机械工程学报, 2026, 62(5): 117-132. |
| [11] | 张涛, 何家豪, 施永平, 赵萍, 汪键, 王伟. 基于改进Stribeck摩擦的协作机器人动力学参数辨识[J]. 机械工程学报, 2026, 62(5): 151-167. |
| [12] | 林松, 戴建生, 宋屹峰, 王洪光, 袁兵兵. 具有可重构躯干的轮式移动机器人曲面运动特性分析[J]. 机械工程学报, 2026, 62(5): 168-181. |
| [13] | 孙鹏, 葛梦虎, 芮超, 曹利康, 陈波, 王剑斌, 李研彪. 混联双足机械腿的动力学建模与参数辨识[J]. 机械工程学报, 2026, 62(5): 192-203. |
| [14] | 郭世航, 王神龙, 侯锦辉, 李永歌. 单磁驱动多自由度软体机器鱼设计与动力学建模[J]. 机械工程学报, 2026, 62(5): 215-229. |
| [15] | 潘雪婷, 杨飞, 耿中泰, 岳洪浩, 徐金随, 王哲平. 柔性机器人刚度调制策略研究综述[J]. 机械工程学报, 2026, 62(3): 311-339. |
| 阅读次数 | ||||||
|
全文 |
|
|||||
|
摘要 |
|
|||||
