Search

Your search keyword '"Gao, Pan"' showing total 1,166 results
Start Over Author "Gao, Pan" Publication Year Range Last 3 years
1,166 results on '"Gao, Pan"'

57. Self-Ensemling for 3D Point Cloud Domain Adaption

Author

Li, Qing, Peng, Xiaojiang, Yan, Chuan, Gao, Pan, and Hao, Qi

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Recently 3D point cloud learning has been a hot topic in computer vision and autonomous driving. Due to the fact that it is difficult to manually annotate a qualitative large-scale 3D point cloud dataset, unsupervised domain adaptation (UDA) is popular in 3D point cloud learning which aims to transfer the learned knowledge from the labeled source domain to the unlabeled target domain. However, the generalization and reconstruction errors caused by domain shift with simply-learned model are inevitable which substantially hinder the model's capability from learning good representations. To address these issues, we propose an end-to-end self-ensembling network (SEN) for 3D point cloud domain adaption tasks. Generally, our SEN resorts to the advantages of Mean Teacher and semi-supervised learning, and introduces a soft classification loss and a consistency loss, aiming to achieve consistent generalization and accurate reconstruction. In SEN, a student network is kept in a collaborative manner with supervised learning and self-supervised learning, and a teacher network conducts temporal consistency to learn useful representations and ensure the quality of point clouds reconstruction. Extensive experiments on several 3D point cloud UDA benchmarks show that our SEN outperforms the state-of-the-art methods on both classification and segmentation tasks. Moreover, further analysis demonstrates that our SEN also achieves better reconstruction results.

Published
2021

58. Patch-Based Deep Autoencoder for Point Cloud Geometry Compression

Author

You, Kang and Gao, Pan

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Multimedia, Electrical Engineering and Systems Science - Image and Video Processing
Abstract

The ever-increasing 3D application makes the point cloud compression unprecedentedly important and needed. In this paper, we propose a patch-based compression process using deep learning, focusing on the lossy point cloud geometry compression. Unlike existing point cloud compression networks, which apply feature extraction and reconstruction on the entire point cloud, we divide the point cloud into patches and compress each patch independently. In the decoding process, we finally assemble the decompressed patches into a complete point cloud. In addition, we train our network by a patch-to-patch criterion, i.e., use the local reconstruction loss for optimization, to approximate the global reconstruction optimality. Our method outperforms the state-of-the-art in terms of rate-distortion performance, especially at low bitrates. Moreover, the compression process we proposed can guarantee to generate the same number of points as the input. The network model of this method can be easily applied to other point cloud reconstruction problems, such as upsampling., Comment: Accepted to ACM Multimedia Asia (MMAsia '21)

Published
2021
Full Text
View/download PDF

75. Prognosis of acute tandem vertebrobasilar artery occlusion after endovascular revascularization: report of 647 cases

Author

GAO Pan, LI Fengli, ZI Wenjie, and YANG Qingwu

Subjects
endovascular therapy, basilar artery, tandem occlusion, good prognosis, mortality, Medicine (General), R5-920
Abstract

Objective To analyze the prognosis and risk factors of acute tandem vertebrobasilar artery occlusion (VBO) after endovascular therapy (EVT). Methods This study used the patient data from the EVT for Acute Basilar Artery Occlusion Study (BASILAR), a national multicenter clinical registry study for endovascular treatment of acute basilar artery occlusion from January 2014 to May 2019. All patients received intravascular treatment and underwent a 90 d follow-up. According to the presence of vertebral artery and basilar artery occlusion, they were divided into vertebral basilar artery occlusion (VBO) group and acute basilar artery occlusion (BAO) group. Their baseline indicators were analyzed to explore the differences in prognosis and influencing factors between the 2 groups. Results A total of 647 eligible patients were included in this study, with a median age of 64 (56, 74) years, and 74.7% of them (483) were males. Among them, 560 patients had simple basilar artery occlusion and the other 87 had basilar artery occlusion combined with severe stenosis or occlusion of the vertebral artery. There were no significant differences in the 90-day good prognosis rate (defined as mRS score of 0~3) (P=0.090) and mortality rate (P=0.890) between the VBO group and the BAO group. For the patients in the VBO group, lower NIHSS score (OR=0.898, 95% CI: 0.845~0.954, P < 0.001), higher pc-ASPECTS score (OR=1.690, 95%CI: 1.117~2.557, P=0.010), and shorter puncture to recanalization time (OR=0.990, 95%CI: 0.980~1.000, P=0.047) were independent risk factors for 90-day good prognosis. Similarly, higher NIHSS score (OR=1.094, 95%CI: 1.033~1.159, P=0.002), lower pc-ASPECTS score (OR=0.622, 95%CI: 0.438~0.882, P=0.008) and longer puncture to recanalization time (OR=1.010, 95%CI: 1.001~1.019, P=0.040) were also predictive factors for 90-day mortality in these patients. Conclusion EVT is safe and effective for acute VBO, with a good short-term prognosis. NIHSS score, pc-ASPECTS score, and puncture to recanalization time are independent predictors for 90-day good prognosis in VBO patients, and are also predictors of 90-day mortality rate.

Published
2023
Full Text
View/download PDF

76. m6A methylation in cellular senescence of age-associated diseases

Author

Gao Pan, Yao Feng, Pang Jin, Yin Kai, and Zhu Xiao

Subjects
cellular senescence, N6-methyladenosine, age-associated diseases, Biochemistry, QD415-436, Genetics, QH426-470
Abstract

Cellular senescence is a state of irreversible cellular growth arrest that occurs in response to various stresses. In addition to exiting the cell cycle, senescent cells undergo many phenotypic alterations, including metabolic reprogramming, chromatin rearrangement, and senescence-associated secretory phenotype (SASP) development. Furthermore, senescent cells can affect most physiological and pathological processes, such as physiological development; tissue homeostasis; tumour regression; and age-associated disease progression, including diabetes, atherosclerosis, Alzheimer’s disease, and hypertension. Although corresponding anti-senescence therapies are actively being explored for the treatment of age-associated diseases, the specific regulatory mechanisms of senescence remain unclear. N6-methyladenosine (m6A), a chemical modification commonly distributed in eukaryotic RNA, plays an important role in biological processes such as translation, shearing, and RNA transcription. Numerous studies have shown that m6A plays an important regulatory role in cellular senescence and aging-related disease. In this review, we systematically summarize the role of m6A modifications in cellular senescence with regard to oxidative stress, DNA damage, telomere alterations, and SASP development. Additionally, diabetes, atherosclerosis, and Alzheimer’s disease regulation via m6A-mediated cellular senescence is discussed. We further discuss the challenges and prospects of m6A in cellular senescence and age-associated diseases with the aim of providing rational strategies for the treatment of these age-associated diseases.

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results