Zobrazeno 1 - 10
of 34
pro vyhledávání: '"Fayao Liu"'
LiDAR-based 3D scene perception is a fundamental and important task for autonomous driving. Most state-of-the-art methods on LiDAR-based 3D recognition tasks focus on single frame 3D point cloud data, and the temporal information is ignored in those
Externí odkaz:
http://arxiv.org/abs/2207.04673
Publikováno v:
IEEE Transactions on Pattern Analysis and Machine Intelligence. 45:6807-6819
Embodied Question Answering (EQA) is a newly defined research area where an agent is required to answer the users questions by exploring the real-world environment. It has attracted increasing research interests due to its broad applications in perso
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::6d9bba82880f8b3b0b248981e31a65de
https://doi.org/10.1109/tpami.2021.3139957
https://doi.org/10.1109/tpami.2021.3139957
Publikováno v:
IEEE Transactions on Multimedia. :1-12
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::fbe46b65c268d1baf35a2839232ce89f
https://doi.org/10.1109/tmm.2023.3262973
https://doi.org/10.1109/tmm.2023.3262973
Publikováno v:
International Journal of Computer Vision. 130:3140-3157
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::69943e95a941090b9f5ad5b177e99de0
https://doi.org/10.1007/s11263-022-01677-7
https://doi.org/10.1007/s11263-022-01677-7
Publikováno v:
IEEE Transactions on Neural Networks and Learning Systems. :1-12
Knowledge distillation is a learning paradigm for boosting resource-efficient graph neural networks (GNNs) using more expressive yet cumbersome teacher models. Past work on distillation for GNNs proposed the Local Structure Preserving loss (LSP), whi
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::a33e7911455786fd9a42f689917513d5
https://doi.org/10.1109/tnnls.2022.3223018
https://doi.org/10.1109/tnnls.2022.3223018
Current vision language pretraining models are dominated by methods using region visual features extracted from object detectors. Given their good performance, the extract-then-process pipeline significantly restricts the inference speed and therefor
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::6be80b49d0d6c225dbf1bddd7852c1d7
http://arxiv.org/abs/2301.07236
http://arxiv.org/abs/2301.07236
The goal of 3D pose transfer is to transfer the pose from the source mesh to the target mesh while preserving the identity information (e.g., face, body shape) of the target mesh. Deep learning-based methods improved the efficiency and performance of
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::35e4ef039402228840095d3ee7b7eb53
http://arxiv.org/abs/2211.10278
http://arxiv.org/abs/2211.10278
Publikováno v:
2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::4760d691c693e6c3e468867226e56f23
https://doi.org/10.1109/cvpr52688.2022.01154
https://doi.org/10.1109/cvpr52688.2022.01154
Natural language BERTs are trained with language corpus in a self-supervised manner. Unlike natural language BERTs, vision language BERTs need paired data to train, which restricts the scale of VL-BERT pretraining. We propose a self-training approach
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::f7bcb60497cc8c95d547c4dc08dfbd35
http://arxiv.org/abs/2201.02010
http://arxiv.org/abs/2201.02010
Publikováno v:
CVPR
Scene flow in 3D point clouds plays an important role in understanding dynamic environments. Although significant advances have been made by deep neural networks, the performance is far from satisfactory as only per-point translational motion is cons
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::5b0fb00ebe9e643abf56bda62825f382