A5007612908- 3D Shape Modeling and Analysis
- Image Retrieval and Classification Techniques
- Advanced Image and Video Retrieval Techniques
- Image Processing and 3D Reconstruction
- Face recognition and analysis
- Generative Adversarial Networks and Image Synthesis
- Advanced Vision and Imaging
- Computer Graphics and Visualization Techniques
- Medical Image Segmentation Techniques
- Robotics and Sensor-Based Localization
- Domain Adaptation and Few-Shot Learning
- 3D Surveying and Cultural Heritage
- Biometric Identification and Security
- Face and Expression Recognition
- Video Surveillance and Tracking Methods
- Digital Media Forensic Detection
- Target Tracking and Data Fusion in Sensor Networks
- Multimodal Machine Learning Applications
- Advanced Image Processing Techniques
- Advanced Neural Network Applications
- Face Recognition and Perception
- Anatomy and Medical Technology
- Antimicrobial Peptides and Activities
- Biological and pharmacological studies of plants
- RNA and protein synthesis mechanisms
Hangzhou Dianzi University
2017-2025
Osaka University
2020
China University of Petroleum, East China
2013-2017
University of North Carolina at Charlotte
2016-2017
China University of Petroleum, Beijing
2013
In this paper, a hierarchical deep multi-task learning (HD-MTL) algorithm is developed to support large-scale visual recognition (e.g., recognizing thousands or even tens of atomic object classes automatically). To achieve more effective accomplishment the coarse-to-fine tasks for recognition, multiple sets features are first extracted from different layers convolutional neural networks (deep CNNs). A tree then learned by assigning visually-similar with similar complexities into same group,...
3D Gaussian Splatting has emerged as one of the most prominent algorithms for novel view synthesis in recent years, with numerous studies adapting it to dynamic scenes, such employing deformation MLP predict motion. However, existing methods frequently overlook contextual information within temporal sequences, resulting inaccuracies motion modeling. To mitigate this issue, we propose KF-GS, which is inspired by Kalman filter algorithm and integrates both observation prediction estimate...
The classic subgroup A (ALV-A) and newly emerging K (ALV-K) of avian leukosis virus are two major pathogens responsible for leukemia in chickens, posing substantial threats to global poultry industry. Both viruses share a Tva protein encoded by the tva gene as receptor gain entry into host cells. In this study, we described identifications alleles Qingyuan partridge chicken, which possesses an 11-nucleotide (GCTGCCCACCC) insertion 6-nucleotide (ACCTCC) deletion independently located exon 1...
The growing application of face images and modern AI technology has raised another important concern in privacy protection. In many real scenarios like scientific research, social sharing commercial application, lots are released without processing to protect people's identity. this paper, we develop a novel effective de-identification generative adversarial network (DeIdGAN) for anonymization by seamlessly replacing given image with different synthesized yet realistic one. Our approach...
To achieve more effective solution for large-scale image classification (i.e., classifying millions of images into thousands or even tens object classes categories), a deep multi-task learning algorithm is developed by seamlessly integrating CNNs with over the concept ontology, where ontology used to organize large numbers categories hierarchically and determine inter-related tasks automatically. Our can integrate learn discriminative high-level features representation, it also leverage...
The applications of isometric 3-D objects have recently received sufficient attention and, thus, it is very attractive to retrieve such from large-scale collections. Although existing approaches presented some interesting ideas, their performance limited ability on feature representation. To improve the object (shape) recognition, recent algorithms prefer using complicated deep neural networks learn discriminative features, but they consume huge amounts computing resources. Instead, this...
Invariance against rotations of 3D objects is an important property in analyzing point set data. Conventional DNNs having rotation invariance typically obtain accurate shape features via supervised learning by using labeled sets as training samples. However, due to the rapid increase data and high cost labeling, a framework learn rotation-invariant from numerous unlabeled required. This paper proposes novel self-supervised for acquiring at object-level. Our proposed lightweight DNN...
In this paper, a deep mixture of diverse experts algorithm is developed to achieve more efficient learning huge (mixture) network for large-scale visual recognition application. First, two-layer ontology constructed assign large numbers atomic object classes into set task groups according the similarities their complexities, where certain degrees inter-group overlapping are allowed enable sufficient message passing. Second, one particular base CNNs with M+1 outputs learned each group...
We present an automatic approach for large-scale fashion recognition, given image without any kind of annotation. formulate the problem as a hierarchical deep learning (HDL) algorithm which can: (i) integrate CNNs to learn more discriminative high-level features representations both coarse-grained and fine-grained classes at different levels ontology tree; (ii) leverage multi-task inter-task relationship constraint train classifiers nodes on ontology; (iii) use back propagation...
With the increasing popularity of 3D models, retrieving deformable objects is becoming a crucial task. The state-of-the-art methods use complex deep neural networks to address this problem, which require lots computational resources. In paper, we develop more effective solution by using point convolution. Our algorithm takes local descriptors as input and produces global vector for shape retrieval. To save efforts designing convolutional network (CNN), first intrinsic describe deformations....