A5016938698- Robotics and Sensor-Based Localization
- Advanced Vision and Imaging
- Advanced Image and Video Retrieval Techniques
- Optical measurement and interference techniques
- 3D Surveying and Cultural Heritage
- Radiomics and Machine Learning in Medical Imaging
- Advanced Image Processing Techniques
- Advanced Neural Network Applications
- Remote Sensing and LiDAR Applications
- 3D Shape Modeling and Analysis
- Image Processing Techniques and Applications
- Image Enhancement Techniques
- Multimodal Machine Learning Applications
- Face and Expression Recognition
- AI in cancer detection
- Pancreatic and Hepatic Oncology Research
- Medical Image Segmentation Techniques
- Brain Tumor Detection and Classification
- Human Pose and Action Recognition
- Inflammatory Bowel Disease
- COVID-19 diagnosis using AI
- Computer Graphics and Visualization Techniques
- Advanced Text Analysis Techniques
- Collaboration in agile enterprises
- Nanocomposite Films for Food Packaging
Gannan Medical University
2025
Dongguan University of Technology
2024
Shenzhen University
2020-2024
University of Hong Kong
2010-2022
Hong Kong University of Science and Technology
2010-2022
Apple (United States)
2022
Shenzhen University Health Science Center
2022
Kunming University of Science and Technology
2021-2022
Sichuan Academy Of Social Sciences
2012
Deep learning has recently demonstrated its excellent performance for multi-view stereo (MVS). However, one major limitation of current learned MVS approaches is the scalability: memory-consuming cost volume regularization makes hard to be applied high-resolution scenes. In this paper, we introduce a scalable framework based on recurrent neural network. Instead regularizing entire 3D in go, proposed Recurrent Multi-view Stereo Network (R-MVSNet) sequentially regularizes 2D maps along depth...
A successful point cloud registration often lies on robust establishment of sparse matches through discriminative 3D local features. Despite the fast evolution learning-based feature descriptors, little attention has been drawn to learning detectors, even less for a joint two tasks. In this paper, we leverage fully convolutional network clouds, and propose novel practical mechanism that densely predicts both detection score description each point. particular, keypoint selection strategy...
While deep learning has recently achieved great success on multi-view stereo (MVS), limited training data makes the trained model hard to be generalized unseen scenarios. Compared with other computer vision tasks, it is rather difficult collect a large-scale MVS dataset as requires expensive active scanners and labor-intensive process obtain ground truth 3D structures. In this paper, we introduce BlendedMVS, novel dataset, provide sufficient for learning-based MVS. To create apply...
Establishing correspondences between two images requires both local and global spatial context. Given putative of feature points in views, this paper, we propose Order-Aware Network, which infers the probabilities being inliers regresses relative pose encoded by essential matrix. Specifically, proposed network is built hierarchically comprises three novel operations. First, to capture context sparse correspondences, clusters unordered input learning a soft assignment These are canonical...
This work focuses on mitigating two limitations in the joint learning of local feature detectors and descriptors. First, ability to estimate shape (scale, orientation, etc.) points is often neglected during dense extraction, while shape-awareness crucial acquire stronger geometric invariance. Second, localization accuracy detected keypoints not sufficient reliably recover camera geometry, which has become bottleneck tasks such as 3D reconstruction. In this paper, we present ASLFeat, with...
Removing outlier correspondences is one of the critical steps for successful feature-based point cloud registration. Despite increasing popularity introducing deep learning techniques in this field, spatial consistency, which essentially established by a Euclidean transformation between clouds, has received almost no individual attention existing frameworks. In paper, we present PointDSC, novel neural network that explicitly incorporates consistency pruning correspondences. First, propose...
Most existing studies on learning local features focus the patch-based descriptions of individual keypoints, whereas neglecting spatial relations established from their keypoint locations. In this paper, we go beyond detail representation by introducing context awareness to augment off-the-shelf feature descriptors. Specifically, propose a unified framework that leverages and aggregates cross-modality contextual information, including (i) visual high-level image representation, (ii)...
Matching local features across images is a fundamental problem in computer vision. Targeting towards high accuracy and efficiency, we propose Seeded Graph Network, graph neural network with sparse structure to reduce redundant connectivity learn compact representation. The consists of 1) Seeding Module, which initializes the matching by generating small set reliable matches as seeds. 2) Neural utilizes seed pass messages within/across predicts assignment costs. Three novel operations are...
Accurate relative pose is one of the key components in visual odometry (VO) and simultaneous localization mapping (SLAM). Recently, self-supervised learning framework that jointly optimizes target image depth has attracted attention community. Previous works rely on photometric error generated from depths poses between adjacent frames, which contains large systematic under realistic scenes due to reflective surfaces occlusions. In this paper, we bridge gap geometric loss by introducing...
Temporal camera relocalization estimates the pose with respect to each video frame in sequence, as opposed one-shot which focuses on a still image. Even though time dependency has been taken into account, current temporal methods generally underperform state-of-the-art approaches terms of accuracy. In this work, we improve method by using network architecture that incorporates Kalman filtering (KFNet) for online relocalization. particular, KFNet extends scene coordinate regression problem...
Learning-based multi-view stereo (MVS) methods have demonstrated promising results. However, very few existing networks explicitly take the pixel-wise visibility into consideration, resulting in erroneous cost aggregation from occluded pixels. In this paper, we infer and integrate occlusion information MVS network via matching uncertainty estimation. The pair-wise map is jointly inferred with depth map, which further used as weighting guidance during volume fusion. As such, adverse influence...
Recently, an emerging trend in medical image classification is to combine radiomics framework with deep learning network integrated system. Although this combination efficient some tasks, the learning-based often difficult capture effective representation of lesion regions, and prone face challenge overfitting, leading unreliable features inaccurate results, especially when sizes lesions are small or training dataset small. In addition, these combinations mostly lack feature selection...
Establishing correspondences between two images requires both local and global spatial context. Given putative of feature points in views, this paper, we propose Order-Aware Network, which infers the probabilities being inliers regresses relative pose encoded by essential matrix. Specifically, proposed network is built hierarchically comprises three novel operations. First, to capture context sparse correspondences, clusters unordered input learning a soft assignment These are canonical...
Establishing correct correspondences between two images should consider both local and global spatial context. Given putative of feature points in views, this paper, we propose Order-Aware Network, which infers the probabilities being inliers regresses relative pose encoded by essential or fundamental matrix. Specifically, proposed network is built hierarchically comprises three operations. First, to capture context sparse correspondences, clusters unordered input learning a soft assignment...