A5047270197- AI in cancer detection
- Robot Manipulation and Learning
- Radiomics and Machine Learning in Medical Imaging
- Medical Image Segmentation Techniques
- Human Pose and Action Recognition
- COVID-19 diagnosis using AI
- Advanced Neural Network Applications
- Robotics and Sensor-Based Localization
- Advanced Vision and Imaging
- Anomaly Detection Techniques and Applications
- Digital Imaging for Blood Diseases
- Robotic Path Planning Algorithms
- Brain Tumor Detection and Classification
- Autonomous Vehicle Technology and Safety
- Domain Adaptation and Few-Shot Learning
- Multimodal Machine Learning Applications
- Image Processing Techniques and Applications
- Hand Gesture Recognition Systems
- Digital Media Forensic Detection
- Network Security and Intrusion Detection
- Medical Imaging and Analysis
- Industrial Vision Systems and Defect Detection
- Advanced Steganography and Watermarking Techniques
- Video Surveillance and Tracking Methods
- 3D Shape Modeling and Analysis
North China University of Technology
2025
Xiamen University of Technology
2025
Guangdong Pharmaceutical University
2025
Shandong University
2024
Chinese Academy of Medical Sciences & Peking Union Medical College
2024
Software (Spain)
2024
Shanghai Electric (China)
2023-2024
Shenzhen Maternity and Child Healthcare Hospital
2024
University of Leicester
2019-2023
Amazon (United States)
2023
Estimating the 6D pose of known objects is important for robots to interact with real world.The problem challenging due variety as well complexity a scene caused by clutter and occlusions between objects.In this work, we introduce PoseCNN, new Convolutional Neural Network object estimation.PoseCNN estimates 3D translation an localizing its center in image predicting distance from camera.The rotation estimated regressing quaternion representation.We also novel loss function that enables...
We introduce DexYCB, a new dataset for capturing hand grasping of objects. first compare DexYCB with related one through cross-dataset evaluation. then present thorough benchmark state-of-the-art approaches on three relevant tasks: 2D object and keypoint detection, 6D pose estimation, 3D estimation. Finally, we evaluate robotics-relevant task: generating safe robot grasps in human-to-robot handover. <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup>
Automatic Non-rigid Histological Image Registration (ANHIR) challenge was organized to compare the performance of image registration algorithms on several kinds microscopy histology images in a fair and independent manner. We have assembled 8 datasets, containing 355 with 18 different stains, resulting 481 pairs be registered. accuracy evaluated using manually placed landmarks. In total, 256 teams registered for challenge, 10 submitted results, 6 participated workshop. Here, we present...
Current 6D object pose estimation methods usually require a 3D model for each object. These also additional training in order to incorporate new objects. As result, they are difficult scale large number of objects and cannot be directly applied unseen We propose novel framework present network that reconstructs latent representation an using small reference views at inference time. Our is able render the from arbitrary views. Using this neural renderer, we optimize given input image. By our...
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">(Aim)</i> Breast cancer is the most common in women and second worldwide. With rapid advancement of deep learning, early stages breast development can be accurately detected by radiologists with help artificial intelligence systems. xmlns:xlink="http://www.w3.org/1999/xlink">(Method)</i> Based on mammographic imaging, a mainstream clinical screening technique, we present diagnostic system for...
Alzheimer’s and related diseases are significant health issues of this era. The interdisciplinary use deep learning in field has shown great promise gathered considerable interest. This paper surveys literature to disease, mild cognitive impairment, from 2010 early 2023. We identify the major types unsupervised, supervised, semi-supervised methods developed for various tasks field, including most recent developments, such as application recurrent neural networks, graph-neural generative...
Traditional deep learning methods are suboptimal in classifying ambiguity features, which often arise noisy and hard to predict categories, especially, distinguish semantic scoring. Semantic scoring, depending on logic implement evaluation, inevitably contains fuzzy description misses some concepts, for example, the ambiguous relationship between normal probably always presents unclear boundaries (normal-more likely normal-probably normal). Thus, human error is common when annotating images....
Tracking 6D poses of objects from videos provides rich information to a robot in performing different tasks such as manipulation and navigation. In this work, we formulate the object pose tracking problem Rao-Blackwellized particle filtering framework, where 3D rotation translation an are decoupled. This factorization allows our approach, called PoseRBPF, efficiently estimate along with full distribution over rotation. is achieved by discretizing space fine-grained manner, training...
Breast cancer is one of the common cancers threatening health women while incident rate it quite low in men to contribute a major killer men. Early syndromes breast including micro-calcification, mass, and distortion mammography images can be very helpful for rad iologists make diagnosis at early stage, which means treated or even cured timely thus important. To assist radiologists with diagnosis, we set up computer-aided system decision this paper. We acquired regions interests mammographic...
Majority of the perception methods in robotics require depth information provided by RGB-D cameras. However, standard 3D sensors fail to capture transparent objects due refraction and absorption light. In this paper, we introduce a new approach for completion from single image. Key our is local implicit neural representation built on ray-voxel pairs that allows method generalize unseen achieve fast inference speed. Based representation, present novel frame-work can complete missing given...
Tracking 6D poses of objects from videos provides rich information to a robot in performing different tasks such as manipulation and navigation.In this work, we formulate the object pose tracking problem Rao-Blackwellized particle filtering framework, where 3D rotation translation an are decoupled.This factorization allows our approach, called PoseRBPF efficiently estimate along with full distribution over rotation.This is achieved by discretizing space fine-grained manner, training...
This paper studies a discrete-time major-minor mean field game of stopping where the major player can choose either an optimal control or time. We look for relaxed equilibrium as randomized policy, which is formulated fixed point set-valued mapping, whose existence challenging by direct arguments. To overcome difficulties caused presence player, we propose to study auxiliary problem considering entropy regularization in player's while formulating minor players' problems linear programming...
Abstract Efficient detection of surface defects is primary for ensuring product quality during manufacturing processes. To enhance the performance deep learning‐based methods in practical applications, authors propose Dense‐YOLO, a fast defect network that combines strengths DenseNet and you only look once version 3 (YOLOv3). The design lightweight backbone with improved densely connected blocks, optimising utilisation shallow features while maintaining high speeds. Additionally, refine...
Abstract This paper introduces a novel approach to predicting the maximum power point in Hadoop Electrical Energy mechanism, addressing inaccuracies inherent traditional methods. We propose model leveraging conventional - Back Propagation Neural Network (GA-BPNN) forecast largest load dot. The simulation results from our numerical examples indicate that GA-BPNN-based MPPT control algorithm for mechanism efficiently tracks dot with high precision and stability, outperforming comparative...