Saved in:
Bibliographic Details
Main Authors: Xia, Yan, Shi, Letian, Ding, Zifeng, Henriques, João F., Cremers, Daniel
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2311.15977
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910387699449856
author Xia, Yan
Shi, Letian
Ding, Zifeng
Henriques, João F.
Cremers, Daniel
author_facet Xia, Yan
Shi, Letian
Ding, Zifeng
Henriques, João F.
Cremers, Daniel
contents We tackle the problem of 3D point cloud localization based on a few natural linguistic descriptions and introduce a novel neural network, Text2Loc, that fully interprets the semantic relationship between points and text. Text2Loc follows a coarse-to-fine localization pipeline: text-submap global place recognition, followed by fine localization. In global place recognition, relational dynamics among each textual hint are captured in a hierarchical transformer with max-pooling (HTM), whereas a balance between positive and negative pairs is maintained using text-submap contrastive learning. Moreover, we propose a novel matching-free fine localization method to further refine the location predictions, which completely removes the need for complicated text-instance matching and is lighter, faster, and more accurate than previous methods. Extensive experiments show that Text2Loc improves the localization accuracy by up to $2\times$ over the state-of-the-art on the KITTI360Pose dataset. Our project page is publicly available at \url{https://yan-xia.github.io/projects/text2loc/}.
format Preprint
id arxiv_https___arxiv_org_abs_2311_15977
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Text2Loc: 3D Point Cloud Localization from Natural Language
Xia, Yan
Shi, Letian
Ding, Zifeng
Henriques, João F.
Cremers, Daniel
Computer Vision and Pattern Recognition
We tackle the problem of 3D point cloud localization based on a few natural linguistic descriptions and introduce a novel neural network, Text2Loc, that fully interprets the semantic relationship between points and text. Text2Loc follows a coarse-to-fine localization pipeline: text-submap global place recognition, followed by fine localization. In global place recognition, relational dynamics among each textual hint are captured in a hierarchical transformer with max-pooling (HTM), whereas a balance between positive and negative pairs is maintained using text-submap contrastive learning. Moreover, we propose a novel matching-free fine localization method to further refine the location predictions, which completely removes the need for complicated text-instance matching and is lighter, faster, and more accurate than previous methods. Extensive experiments show that Text2Loc improves the localization accuracy by up to $2\times$ over the state-of-the-art on the KITTI360Pose dataset. Our project page is publicly available at \url{https://yan-xia.github.io/projects/text2loc/}.
title Text2Loc: 3D Point Cloud Localization from Natural Language
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2311.15977