SO-SLAM: Semantic Object SLAM With Scale Proportional and Symmetrical Texture Constraints

Liao, Ziwei; Hu, Yutong; Zhang, Jiadong; Qi, Xianyu; Zhang, Xiaoyu; Wang, Wei

doi:10.1109/lra.2022.3148465

Cited by 40 publications

(18 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Robots should have the ability to have a human-centered understanding of their environment. It needs to distinguish between a room and a hallway, or the different functions of a kitchen and a living room in the future [232]. Therefore, semantic attributes involving human concepts (such as room types, objects, and their spatial layout), which is considered a necessary attribute of future robots [233].…”

Section: Semantic With Mappingmentioning

confidence: 99%

An Overview on Visual SLAM: From Tradition to Semantic

Chen

Shang

et al. 2022

Remote Sensing

View full text Add to dashboard Cite

Visual SLAM (VSLAM) has been developing rapidly due to its advantages of low-cost sensors, the easy fusion of other sensors, and richer environmental information. Traditional visionbased SLAM research has made many achievements, but it may fail to achieve wished results in challenging environments. Deep learning has promoted the development of computer vision, and the combination of deep learning and SLAM has attracted more and more attention. Semantic information, as high-level environmental information, can enable robots to better understand the surrounding environment. This paper introduces the development of VSLAM technology from two aspects: traditional VSLAM and semantic VSLAM combined with deep learning. For traditional VSLAM, we summarize the advantages and disadvantages of indirect and direct methods in detail and give some classical VSLAM open-source algorithms. In addition, we focus on the development of semantic VSLAM based on deep learning. Starting with typical neural networks CNN and RNN, we summarize the improvement of neural networks for the VSLAM system in detail. Later, we focus on the help of target detection and semantic segmentation for VSLAM semantic information introduction. We believe that the development of the future intelligent era cannot be without the help of semantic technology. Introducing deep learning into the VSLAM system to provide semantic information can help robots better perceive the surrounding environment and provide people with higher-level help.

show abstract

Section: Semantic With Mappingmentioning

confidence: 99%

An Overview on Visual SLAM: From Tradition to Semantic

Chen

Shang

et al. 2022

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Their experiments show that incorporating objects and planes as new factors produce semantically meaningful maps and more accurate trajectories. In SO-SLAM [20], Liao et al used manually extracted planes to add supporting constraints to objects, as well as, semantic scale priors and symmetry constraints.…”

Section: Object-based Slammentioning

confidence: 99%

Section: Motivationsmentioning

confidence: 99%

“…Most of the existing object-based SLAM systems [14,14,20,25,26,34] integrated objects in the core of a SLAM system through a joint camera-landmark-object optimization. However, [20] and the monocular version of [34] noticed that it did not significantly improve the camera pose accuracy and [25] even observed a decreased accuracy compared to the point-based tracking of ORB-SLAM2. Objects represented with coarse models (cuboids or ellipsoids) might thus not be accurate enough to improve the camera pose tracking, but we nevertheless think that they are of great interest when the tracking gets lost.…”

Section: Motivationsmentioning

confidence: 99%

“…Indeed, most existing methods require external information or manual processing. For example, QuadricSLAM [25] and SO-SLAM [20] assume that the problem of object association is solved. In their experiments, they manually annotated matches between the detected objects and the objects in the map.…”

Section: Motivationsmentioning

confidence: 99%

See 2 more Smart Citations