A5100698922- Generative Adversarial Networks and Image Synthesis
- Immune Response and Inflammation
- Immune cells in cancer
- Advanced Neural Network Applications
- T-cell and B-cell Immunology
- Visual Attention and Saliency Detection
- Advanced Image and Video Retrieval Techniques
- Cell Image Analysis Techniques
- AI in cancer detection
- Immune responses and vaccinations
- Image and Video Quality Assessment
- Domain Adaptation and Few-Shot Learning
- Face recognition and analysis
- Machine Learning in Healthcare
- Bladder and Urothelial Cancer Treatments
- Bayesian Modeling and Causal Inference
- Probiotics and Fermented Foods
- Renal and related cancers
- Usability and User Interface Design
- Robotics and Sensor-Based Localization
- Autonomous Vehicle Technology and Safety
- Frailty in Older Adults
- Medical Research and Treatments
- Renal cell carcinoma treatment
- Antimicrobial Peptides and Activities
Chinese PLA General Hospital
2023
University of Edinburgh
2019-2022
The University of Sydney
2021
Harbin University of Science and Technology
2020
Anhui University
2019
Tianjin University
2019
Nanyang Technological University
2011
Northeastern University
2003
This paper aims at high-accuracy 3D object detection in autonomous driving scenario. We propose Multi-View networks (MV3D), a sensory-fusion framework that takes both LIDAR point cloud and RGB images as input predicts oriented bounding boxes. encode the sparse with compact multi-view representation. The network is composed of two subnetworks: one for proposal generation another feature fusion. generates candidate boxes efficiently from birds eye view representation cloud. design deep fusion...
Image saliency detection is an active research topic in the community of computer vision and multimedia. Fusing complementary RGB thermal infrared data has been proven to be effective for image detection. In this paper, we propose approach RGB-T Our relies on a novel collaborative graph learning algorithm. particular, take superpixels as nodes, collaboratively use hierarchical deep features jointly learn affinity node unified optimization framework. Moreover, contribute more challenging...
RGB-Thermal saliency detection is to use thermal infrared information assist salient object with visible light information. Multi-Modal Multi-Scale Noise-Insensitive Ranking (M3S-NIR), proposed for (RGB-T) detection. Given spatially aligned RGB and images, M3S-NIR first segments them together into a set of multi-scale superpixels. Second, it takes these superpixels as graph nodes performs multi-modal manifold ranking achieve calculation, in which the cross-modal cross-scale collaborations...
Age has important implications for health, and understanding how age manifests in the human body is first step a potential intervention. This becomes especially cardiac since main risk factor development of cardiovascular disease. Data-driven modeling progression been conducted successfully diverse applications such as face or brain aging. While longitudinal data preferred option training deep learning models, collecting dataset usually very costly, medical imaging. In this work, conditional...
Pseudo healthy synthesis, i.e. the creation of a subject-specific `healthy' image from pathological one, could be helpful in tasks such as anomaly detection, understanding changes induced by pathology and disease or even data augmentation. We treat this task factor decomposition problem: we aim to separate what appears where is (as map). The two factors are then recombined (by network) reconstruct input image. train our models an adversarial way using either paired unpaired settings, pair...
Cone-Beam Computed Tomography (CBCT) imaging modality is used to acquire 3D volumetric image of the human body. CBCT plays a vital role in diagnosing dental diseases, especially cyst or tumour-like lesions. Current computer-aided detection and diagnostic systems have demonstrated value range however, capability such deep learning method on transmissive lesions has not been investigated. In this study, we propose an automatic for jawbones using images. We integrated pre-trained DenseNet with...
We propose in this paper an enhanced age simulation method for face recognition across differences. Since the within-class variations caused by differences are generally much larger than between-class different identities at similar ages, we to reduce facial appearance difference of same person variations. it's very difficult obtain a sequence images continuous first implement filling process enlarge our limited training database and classify it into groups. Then, represent each group mean...
Objective To investigate the incidence of middle ear barotrauma due to hybaric oxygen therapy by using Health Education Atlas . Methods 100 patients were divided into two groups random number table.The research group(49 patients) was educated Atlas. The control group(51 traditional education approach.During first three days, we observed and recorded eardrum injury asked discomfort everyday. Results The rate group 6.1%, which 19.6%.The milder than group(χ2=4.02,...