From Comparison to Retrieval: Scalable Change Retrieval from Discriminatively Learned Deep Three-dimensional Neural Codes

Yusuke, Kojima; Kanji, Tanaka; Naiming, Yang; Hirota, Yuji; Yamaguchi, Kousuke

doi:10.1109/itsc.2019.8917170

Cited by 2 publications

(1 citation statement)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Some researchers segmented the 3D point clouds, comparing each part of it with parts in the retrieved 3D point clouds to generate their localization [12]. Other researchers unified the coordinate system of 3D point clouds, which were input into the neural network for training and localization of prediction or the extracted feature for comparative analysis [13]. In some studies, 3D point clouds were converted into 2D images for analysis.…”

Section: Introduction (Heading 1)mentioning

confidence: 99%

Robot visual place recognition using COG_SC descriptor

YE,

LI,

GUO

et al. 2024

Preprint

View full text Add to dashboard Cite

One of the most active research areas in robotics is visual place recognition using a 3D laser Lidar. A promising research direction involves learning and recognition of scene descriptors, such as scan context descriptors. These descriptors map 3D point clouds to 2D point clouds. Although the scan-context descriptor has a high recognition performance, it is still expensive image data and cannot be handled with low-capacity non-deep models. To address this, we have developed a novel feature descriptor by combining COG (center of gravity) with scan-context descriptor.

show abstract

Section: Introduction (Heading 1)mentioning

confidence: 99%

Robot visual place recognition using COG_SC descriptor

YE,

LI,

GUO

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

Improved Visual Robot Place Recognition of Scan-Context Descriptors by Combining with CNN and SVM

Ye,

Tanaka

2023

J. Robot. Mechatron.

View full text Add to dashboard Cite

Visual place recognition from a 3D laser LiDAR is one of the most active research areas in robotics. Especially, learning and recognition of scene descriptors, such as scan context descriptors that map 3D point clouds to 2D point clouds, is one of the promising research directions. Although the scan-context descriptor has a sufficiently high recognition performance, it is still expensive image data and cannot be handled with low-capacity non-deep models. In this paper, we explore the task of compressing the scan context descriptor model while maintaining its recognition performance. To this end, the proposed approach slightly modifies the off-the-shelf classifier model of convolutional neural networks (CNN) from its basis, by replacing the SoftMax part with a support vector machine (SVM). Experiments with publicly available NCLT dataset validate the effectiveness of the proposed approach.

show abstract

From Comparison to Retrieval: Scalable Change Retrieval from Discriminatively Learned Deep Three-dimensional Neural Codes

Cited by 2 publications

References 38 publications

Robot visual place recognition using COG_SC descriptor

Robot visual place recognition using COG_SC descriptor

Improved Visual Robot Place Recognition of Scan-Context Descriptors by Combining with CNN and SVM

Contact Info

Product

Resources

About