Filters








12,816 Hits in 6.6 sec

Leveraging Temporal Information for 3D Detection and Domain Adaptation [article]

Cunjun Yu, Zhongang Cai, Daxuan Ren, Haiyu Zhao
2020 arXiv   pre-print
However, recent progress largely focuses on detecting objects in a single 360-degree sweep, without extensively exploring the temporal information.  ...  In this report, we describe a simple way to pass such information in the learning pipeline by adding timestamps to the point clouds, which shows consistent improvements across all three classes.  ...  Conclusion We show in this report that temporal information is helpful to detection and in turn, beneficial to domain adaptation on the point clouds.  ... 
arXiv:2006.16796v1 fatcat:xrjcjlyhyjg6nik3qor3h4uvhy

FAST3D: Flow-Aware Self-Training for 3D Object Detectors [article]

Christian Fruhwirth-Reisinger, Michael Opitz, Horst Possegger, Horst Bischof
2021 arXiv   pre-print
To address this issue, we propose a flow-aware self-training method that enables unsupervised domain adaptation for 3D object detectors on continuous LiDAR point clouds.  ...  In order to get reliable pseudo-labels, we leverage scene flow to propagate detections through time.  ...  Acknowledgements The financial support by the Austrian Federal Ministry for Digital and Economic Affairs, the National Foundation for Research, Technology and Development and the Christian Doppler Research  ... 
arXiv:2110.09355v1 fatcat:26y4ehaejvg4lhwvij7fpqgdda

3D Object Detection for Autonomous Driving: A Review and New Outlooks [article]

Jiageng Mao, Shaoshuai Shi, Xiaogang Wang, Hongsheng Li
2022 arXiv   pre-print
This paper reviews the advances in 3D object detection for autonomous driving. First, we introduce the background of 3D object detection and discuss the challenges in this task.  ...  and planning. 3D object detection, which intelligently predicts the locations, sizes, and categories of the critical 3D objects near an autonomous vehicle, is an important part of a perception system.  ...  domain adaptation milestones for 3D object detection based on transferred domains and techniques.  ... 
arXiv:2206.09474v1 fatcat:3skws77uqngjtpo6mycpo4dhny

SF-UDA^3D: Source-Free Unsupervised Domain Adaptation for LiDAR-Based 3D Object Detection [article]

Cristiano Saltori, Stéphane Lathuiliére, Nicu Sebe, Elisa Ricci, Fabio Galasso
2020 arXiv   pre-print
This paper proposes SF-UDA^3D, the first Source-Free Unsupervised Domain Adaptation (SF-UDA) framework to domain-adapt the state-of-the-art PointRCNN 3D detector to target domains for which we have no  ...  SF-UDA^3D outperforms both previous domain adaptation techniques based on features alignment and state-of-the-art 3D object detection methods which additionally use few-shot target annotations or target  ...  We thank the CARITRO Deep Learning laboratory of ProM Facility for the granted GPU time.  ... 
arXiv:2010.08243v2 fatcat:4kbapwfjm5gv7phpekgexnnvem

Unsupervised Domain Adaptive 3D Detection with Multi-Level Consistency [article]

Zhipeng Luo, Zhongang Cai, Changqing Zhou, Gongjie Zhang, Haiyu Zhao, Shuai Yi, Shijian Lu, Hongsheng Li, Shanghang Zhang, Ziwei Liu
2021 arXiv   pre-print
underlying factors of the domain gap in 3D detection.  ...  In addition, existing 3D domain adaptive detection methods often assume prior access to the target domain annotations, which is rarely feasible in the real world.  ...  Acknowledgements This study is supported by NTU NAP, and under the RIE2020 Industry Alignment Fund -Industry Collaboration Projects (IAF-ICP) Funding Initiative, as well as cash and in-kind contribution  ... 
arXiv:2107.11355v2 fatcat:55o6pvwqtndbvgjvjwov63cilq

ST3D++: Denoised Self-training for Unsupervised Domain Adaptation on 3D Object Detection [article]

Jihan Yang, Shaoshuai Shi, Zhe Wang, Hongsheng Li, Xiaojuan Qi
2021 arXiv   pre-print
In this paper, we present a self-training method, named ST3D++, with a holistic pseudo label denoising pipeline for unsupervised domain adaptation on 3D object detection.  ...  Finally, our method is assessed on four 3D benchmark datasets (i.e., Waymo, KITTI, Lyft, and nuScenes) for three common categories (i.e., car, pedestrian and bicycle).  ...  This task is also known as unsupervised domain adaptation (UDA) for 3D object detection.  ... 
arXiv:2108.06682v1 fatcat:nhpe3pvcufeabhtme5ou2fvxxi

Advancing Medical Imaging Informatics by Deep Learning-Based Domain Adaptation

Anirudh Choudhary, Li Tong, Yuanda Zhu, May D. Wang
2020 IMIA Yearbook of Medical Informatics  
Domain adaptation (DA) has been developed to transfer the knowledge from a labeled data domain to a related but unlabeled domain in either image space or feature space.  ...  We aim to summarize recent advances, highlighting the motivation, challenges, and opportunities, and to discuss promising directions for future work in DA for medical imaging.  ...  Leveraging Synthetic Data DA for medical imaging can be applied in relatively under-explored applications such as single-view 3D reconstruction [94] or temporal disease analysis [95] .  ... 
doi:10.1055/s-0040-1702009 pmid:32823306 fatcat:gtlhoh6m3fh4hcumfzdlpdohr4

Self-Supervised Adaptation of High-Fidelity Face Models for Monocular Performance Tracking

Jae Shin Yoon, Takaaki Shiratori, Shoou-I Yu, Hyun Soo Park
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Then, we overcome the domain mismatch between lab and uncontrolled environments by performing self-supervised domain adaptation based on "consecutive frame texture consistency" based on the assumption  ...  However, driving these realistic face models requires special input data, e.g. 3D meshes and unwrapped textures.  ...  Our approach leverages the assumption that the textures (appearance) of a face between consecutive frames should be consistent and incorporates this source of supervision to adapt the domain of I2ZNet  ... 
doi:10.1109/cvpr.2019.00473 dblp:conf/cvpr/YoonSYP19 fatcat:gyih7ownwzbrrkegta6kvwfq34

Offboard 3D Object Detection from Point Cloud Sequences [article]

Charles R. Qi, Yin Zhou, Mahyar Najibi, Pei Sun, Khoa Vo, Boyang Deng, Dragomir Anguelov
2021 arXiv   pre-print
Existing 3D object detectors fail to satisfy the high-quality requirement for offboard uses due to the limited input and speed constraints.  ...  Observing that different frames capture complementary views of objects, we design the offboard detector to make use of the temporal points through both multi-frame object detection and novel object-centric  ...  B presents more evaluation results on the Waymo Open Dataset test set and shows how our offboard 3D detection can help domain adaptation and 3D tracking. Sec.  ... 
arXiv:2103.05073v1 fatcat:xihsnj6ayvdf7ncknb7mxpf5ya

Deep Weakly-Supervised Domain Adaptation for Pain Localization in Videos [article]

Gnana Praveen R, Eric Granger, Patrick Cardinal
2020 arXiv   pre-print
Given the cost of annotating intensity levels for every video frame, we propose a weakly-supervised domain adaptation (WSDA) technique that allows for training 3D CNNs for spatio-temporal pain intensity  ...  The training process relies on weak target loss, along with domain loss and source loss for domain adaptation of the I3D model.  ...  adaptation leverages the wide range of variation of both domains by minimizing the domain differences.  ... 
arXiv:1910.08173v2 fatcat:p7huseb36ve37nk4qpmbygnazq

Self-Supervised Adaptation of High-Fidelity Face Models for Monocular Performance Tracking [article]

Jae Shin Yoon, Takaaki Shiratori, Shoou-I Yu, Hyun Soo Park
2019 arXiv   pre-print
Then, we overcome the domain mismatch between lab and uncontrolled environments by performing self-supervised domain adaptation based on "consecutive frame texture consistency" based on the assumption  ...  However, driving these realistic face models requires special input data, e.g. 3D meshes and unwrapped textures.  ...  Our approach leverages the assumption that the textures (appearance) of a face between consecutive frames should be consistent and incorporates this source of supervision to adapt the domain of I2ZNet  ... 
arXiv:1907.10815v1 fatcat:5kvwovdderetto73mcjki4x67i

Unsupervised Domain Adaptation for Monocular 3D Object Detection via Self-Training [article]

Zhenyu Li, Zehui Chen, Ang Li, Liangji Fang, Qinhong Jiang, Xianming Liu, Junjun Jiang
2022 arXiv   pre-print
Then, we propose STMono3D, a new self-teaching framework for unsupervised domain adaptation on Mono3D.  ...  Monocular 3D object detection (Mono3D) has achieved unprecedented success with the advent of deep learning techniques and emerging large-scale autonomous driving datasets.  ...  We call this task unsupervised domain adaptation (UDA) for monocular 3D object detection.  ... 
arXiv:2204.11590v2 fatcat:oho5oi4go5fxtavavfkqeoiqme

1st Place Solution for Waymo Open Dataset Challenge – 3D Detection and Domain Adaptation [article]

Zhuangzhuang Ding, Yihan Hu, Runzhou Ge, Li Huang, Sijia Chen, Yu Wang, Jie Liao
2020 arXiv   pre-print
In this technical report, we introduce our winning solution "HorizonLiDAR3D" for the 3D detection track and the domain adaptation track in Waymo Open Dataset Challenge at CVPR 2020.  ...  The final detection performance also benefits from model ensemble and Test-Time Augmentation (TTA) in both the 3D detection track and the domain adaptation track.  ...  We ensemble AFDet-B1, AFDet-B2 and AFDet-B3 for 3D detection and AFDet-B1 and AFDet-B2 for domain adaptation in our final submission. Results 3D Detection.  ... 
arXiv:2006.15505v1 fatcat:mmiqzrms5zc5nb7zt45tmuzkfy

Deep Learning for Micro-expression Recognition: A Survey [article]

Yante Li, Jinsheng Wei, Yang Liu, Janne Kauttonen, Guoying Zhao
2021 arXiv   pre-print
Early methods for MER mainly based on traditional appearance and geometry features.  ...  This survey defines a new taxonomy for the field, encompassing all aspects of MER based on DL. For each aspect, the basic approaches and advanced developments are summarized and discussed.  ...  Compared with the static image, the dynamic input is able to leverage spatial and temporal information for robust MER.  ... 
arXiv:2107.02823v4 fatcat:w2bqbvxw4zbc7jjnqmx2gmh5aa

Deep Learning for Face Anti-Spoofing: A Survey [article]

Zitong Yu, Yunxiao Qin, Xiaobai Li, Chenxu Zhao, Zhen Lei, Guoying Zhao
2022 arXiv   pre-print
., pseudo depth map); 2) in addition to traditional intra-dataset evaluation, we collect and analyze the latest methods specially designed for domain generalization and open-set FAS; and 3) besides commercial  ...  It covers several novel and insightful components: 1) besides supervision with binary label (e.g., '0' for bonafide vs. '1' for PAs), we also investigate recent methods with pixel-wise supervision (e.g  ...  On one hand, domain adaptation and generalization techniques are leveraged for robust live/spoof classification under unlimited domain variations.  ... 
arXiv:2106.14948v2 fatcat:wsheo7hbwvewhjoe6ykwjuqfii
« Previous Showing results 1 — 15 out of 12,816 results