A5101922699- Natural Language Processing Techniques
- Remote-Sensing Image Classification
- Pediatric Urology and Nephrology Studies
- Aeolian processes and effects
- Kidney Stones and Urolithiasis Treatments
- Hydrology and Sediment Transport Processes
- Cultural Heritage Materials Analysis
- Artificial Immune Systems Applications
- Minerals Flotation and Separation Techniques
- Generative Adversarial Networks and Image Synthesis
- Image Enhancement Techniques
- Network Security and Intrusion Detection
- Advanced Image Processing Techniques
- Anomaly Detection Techniques and Applications
- Plant Water Relations and Carbon Dynamics
- Advanced Image and Video Retrieval Techniques
- Automated Road and Building Extraction
- Sentiment Analysis and Opinion Mining
- Topic Modeling
- Speech Recognition and Synthesis
- Advanced X-ray and CT Imaging
- Fault Detection and Control Systems
- Wood and Agarwood Research
- Image Retrieval and Classification Techniques
- Digital Media Forensic Detection
The Second Nanning People's Hospital
2023
Northwest University
2023
Guangxi Medical University
2023
Alibaba Group (United States)
2022
Central South University
2021
Heilongjiang University of Science and Technology
2015
Accurate and in-time performance monitoring plays a great role in industrial processes. However, since the labeled performances are usually measured by some special devices with relatively long-time interval, current deep neural networks for always treat them only as an output. But actually, process like froth flotation, at previous moment could also offer valuable information moment, especially under unstable working conditions. Therefore, we propose Siamese time series difference network...
In the scenario of limited labeled remote-sensing datasets, model’s performance is constrained by insufficient availability data. Generative model-based data augmentation has emerged as a promising solution to this limitation. While existing generative models perform well in natural scene domains (e.g., faces and street scenes), their remote sensing hindered severe imbalance semantic similarity among land-cover classes. To tackle these challenges, we propose Multi-Class Guided GAN (MCGGAN),...
The composition of kidney stones is related to the hardness stones. Knowing before surgery can help plan laser power and operation time percutaneous nephroscopic surgery. Moreover, patients be treated with medications if stone compounded by uric acid treatment, which relieve pain However, although literature generally reports analysis method base on dual-energy CT images, accuracy these methods not enough; they need manual delineation location, cannot analyze mixed stones.This study aimed...
Interpreting hierarchical structures latent in language is a key limitation of current models (LMs). While previous research has implicitly leveraged these hierarchies to enhance LMs, approaches for their explicit encoding are yet be explored. To address this, we introduce novel approach re-train transformer encoder-based LMs as Hierarchy Transformer encoders (HiTs), harnessing the expansive nature hyperbolic space. Our method situates output embedding space pre-trained within Poincar\'e...
Virtual try-on (VTON) aims at fitting target clothes to reference person images, which is widely adopted in e-commerce. Existing VTON approaches can be narrowly categorized into Parser-Based (PB) and Parser-Free (PF) by whether relying on the parser information mask persons’clothes synthesize images. Although abandoning has improved applicability of PF methods, ability detail synthesizing also been sacrificed. As a result, distraction from original cloth may persist synthesized especially...
As a crucial computer vision task, multi-objective semantic segmentation has attracted widespread attention and research in the field of remote sensing image analysis. This technology important application value fields such as land resource surveys, global change monitoring, urban planning, environmental monitoring. However, multi-target images faces challenges complex surface features, spectral wide spatial range, resulting differences dimensions among target features. To fully exploit...
Disease labeling is usually the first step to protect and repair mural relics scientifically effectively. The most common method for disease at present manual mapping of regions on orthophoto image by humancomputer interaction. However, this not only time consuming, but also has different result due expert experience. Recently, with development artificial intelligence, machine learning other technologies, automatic can be achieved deep so on. wide variety varying manifestations diseases,...