Boosting Monocular 3D Object Detection With Object-Centric Auxiliary Depth Supervision
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Kim, Youngseok | - |
dc.contributor.author | Kim, Sanmin | - |
dc.contributor.author | Sim, Sangmin | - |
dc.contributor.author | Choi, Jun Won | - |
dc.contributor.author | Kum, Dongsuk | - |
dc.date.accessioned | 2023-05-03T10:21:10Z | - |
dc.date.available | 2023-05-03T10:21:10Z | - |
dc.date.created | 2023-01-05 | - |
dc.date.issued | 2023-02 | - |
dc.identifier.issn | 1524-9050 | - |
dc.identifier.uri | https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/185118 | - |
dc.description.abstract | Recent advances in monocular 3D detection leverage a depth estimation network explicitly as an intermediate stage of the 3D detection network. Depth map approaches yield more accurate depth to objects than other methods thanks to the depth estimation network trained on a large-scale dataset. However, depth map approaches can be limited by the accuracy of the depth map, and sequentially using two separated networks for depth estimation and 3D detection significantly increases computation cost and inference time. In this work, we propose a method to boost the RGB image-based 3D detector by jointly training the detection network with a depth prediction loss analogous to the depth estimation task. In this way, our 3D detection network can be supervised by more depth supervision from raw LiDAR points, which does not require any human annotation cost, to estimate accurate depth without explicitly predicting the depth map. Our novel object-centric depth prediction loss focuses on depth around foreground objects, which is important for 3D object detection, to leverage pixel-wise depth supervision in an object-centric manner. Our depth regression model is further trained to predict the uncertainty of depth to represent the 3D confidence of objects. To effectively train the 3D detector with raw LiDAR points and to enable end-to-end training, we revisit the regression target of 3D objects and design a network architecture. Extensive experiments on KITTI and nuScenes benchmarks show that our method can significantly boost the monocular image-based 3D detector to outperform depth map approaches while maintaining the real-time inference speed. | - |
dc.language | 영어 | - |
dc.language.iso | en | - |
dc.publisher | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC | - |
dc.title | Boosting Monocular 3D Object Detection With Object-Centric Auxiliary Depth Supervision | - |
dc.type | Article | - |
dc.contributor.affiliatedAuthor | Choi, Jun Won | - |
dc.identifier.doi | 10.1109/TITS.2022.3224082 | - |
dc.identifier.scopusid | 2-s2.0-85144084129 | - |
dc.identifier.wosid | 000912789500001 | - |
dc.identifier.bibliographicCitation | IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, v.24, no.2, pp.1801 - 1813 | - |
dc.relation.isPartOf | IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS | - |
dc.citation.title | IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS | - |
dc.citation.volume | 24 | - |
dc.citation.number | 2 | - |
dc.citation.startPage | 1801 | - |
dc.citation.endPage | 1813 | - |
dc.type.rims | ART | - |
dc.type.docType | Article; Early Access | - |
dc.description.journalClass | 1 | - |
dc.description.isOpenAccess | N | - |
dc.description.journalRegisteredClass | scie | - |
dc.description.journalRegisteredClass | scopus | - |
dc.relation.journalResearchArea | Engineering | - |
dc.relation.journalResearchArea | Transportation | - |
dc.relation.journalWebOfScienceCategory | Engineering, Civil | - |
dc.relation.journalWebOfScienceCategory | Engineering, Electrical & Electronic | - |
dc.relation.journalWebOfScienceCategory | Transportation Science & Technology | - |
dc.subject.keywordPlus | Cost benefit analysis | - |
dc.subject.keywordPlus | Cost estimating | - |
dc.subject.keywordPlus | Deep learning | - |
dc.subject.keywordPlus | Forecasting | - |
dc.subject.keywordPlus | Large dataset | - |
dc.subject.keywordPlus | Network architecture | - |
dc.subject.keywordPlus | Object recognition | - |
dc.subject.keywordPlus | Optical radar | - |
dc.subject.keywordPlus | Regression analysis | - |
dc.subject.keywordPlus | Object detection | - |
dc.subject.keywordPlus | 3D object | - |
dc.subject.keywordPlus | 3d object detection | - |
dc.subject.keywordPlus | Autonomous driving | - |
dc.subject.keywordPlus | Auxiliary supervision | - |
dc.subject.keywordPlus | Deep learning | - |
dc.subject.keywordPlus | Depth Estimation | - |
dc.subject.keywordPlus | Depthmap | - |
dc.subject.keywordPlus | Detection networks | - |
dc.subject.keywordPlus | Monocular image | - |
dc.subject.keywordPlus | Objects detection | - |
dc.subject.keywordAuthor | 3D object detection | - |
dc.subject.keywordAuthor | monocular image | - |
dc.subject.keywordAuthor | auxiliary supervision | - |
dc.subject.keywordAuthor | autonomous driving | - |
dc.subject.keywordAuthor | deep learning | - |
dc.identifier.url | https://ieeexplore.ieee.org/document/9966379 | - |
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.
222, Wangsimni-ro, Seongdong-gu, Seoul, 04763, Korea+82-2-2220-1365
COPYRIGHT © 2021 HANYANG UNIVERSITY.
Certain data included herein are derived from the © Web of Science of Clarivate Analytics. All rights reserved.
You may not copy or re-distribute this material in whole or in part without the prior written consent of Clarivate Analytics.