A5078401802- Domain Adaptation and Few-Shot Learning
- Advanced Neural Network Applications
- COVID-19 diagnosis using AI
- Radiomics and Machine Learning in Medical Imaging
- Remote-Sensing Image Classification
- Advanced Image and Video Retrieval Techniques
- Infrastructure Maintenance and Monitoring
- Cancer-related molecular mechanisms research
- Machine Learning and Data Classification
- AI in cancer detection
- Hydrocarbon exploration and reservoir analysis
- Remote Sensing and LiDAR Applications
- Automated Road and Building Extraction
- Multimodal Machine Learning Applications
- Hydraulic Fracturing and Reservoir Analysis
- Seismic Imaging and Inversion Techniques
Universidade Federal de Minas Gerais
2019-2024
Semantic segmentation is a classic computer vision task with multiple applications, which includes medical and remote sensing image analysis. Despite recent advances deep-based approaches, labeling samples (pixels) for training models laborious and, in some cases, unfeasible. In this paper, we present two novel meta-learning methods, named WeaSeL ProtoSeg, the few-shot semantic sparse annotations. We conducted an extensive evaluation of proposed methods different applications (12 datasets)...
It is undeniable that aerial/satellite images can provide useful information for a large variety of tasks. But, since these are always taken from above, some applications benefit complementary provided by other perspective views the scene, such as ground-level images. Despite number public repositories both georeferenced photographs and aerial images, there lack benchmark datasets allow development approaches exploit benefits complementarity aerial/ground imagery. In this article, we present...
The identification of bridges in major infrastructure works is crucial to provide information about the status these constructions and support possible decision-making processes. Typically, this performed by human agents that must detect into large-scale datasets, analyzing image image, a time-consuming task. In paper, we propose novel tool perform bridge detection remote sensing datasets. This implements deep learning-based algorithm, Faster R-CNN (Regions with CNN features), technique...
Soil erosion is considered one of the most expensive natural hazards with a high impact on several infrastructure assets. Among them, railway lines are likely constructions for appearance and, consequently, troublesome due to maintenance costs, risks derailments, and so on. Therefore, it fundamental identify monitor in prevent major consequences. Currently, identification manually performed by humans using huge image sets, time-consuming slow task. Hence, automatic machine learning methods...
Automatic and semi-automatic radiological image segmentation can help physicians in the processing of real-world medical data for several tasks such as detection/diagnosis diseases surgery planning. Current methods based on neural networks are highly data-driven, often requiring hundreds laborious annotations to properly converge. The generalization capabilities traditional supervised deep learning also limited by insufficient variability present training dataset. One very proliferous...
In this paper, we propose a novel approach for few-shot semantic segmentation with sparse labeled images. We investigate the effectiveness of our method, which is based on Model-Agnostic Meta-Learning (MAML) algorithm, in medical scenario, where use labeling and can alleviate cost producing new annotated datasets. Our method uses labels meta-training dense meta-test, thus making model learn to predict from ones. conducted experiments four Chest X-Ray datasets evaluate two types annotations...
Semantic segmentation is a classic computer vision task with multiple applications, which includes medical and remote sensing image analysis. Despite recent advances deep-based approaches, labeling samples (pixels) for training models laborious and, in some cases, unfeasible. In this paper, we present two novel meta learning methods, named WeaSeL ProtoSeg, the few-shot semantic sparse annotations. We conducted extensive evaluation of proposed methods different applications (12 datasets)...
Semantic segmentation is a difficult task in computer vision that have applications many scenarios, often as preprocessing step for tool. Current solutions are based on Deep Neural Networks, which require large amount of data learning task. Aiming to alleviate the strenuous data-collecting/annotating labor, research fields emerged recent years. One them Meta-Learning, tries improve generability models learn restricted data. In this work, we extend previous paper conducting more extensive...
Most uses of Meta-Learning in visual recognition are very often applied to image classification, with a relative lack works other tasks {such} as segmentation and detection. We propose generic framework for few-shot weakly-supervised medical imaging domains. conduct comparative analysis meta-learners from distinct paradigms adapted different sparsely annotated radiological tasks. The modalities include 2D chest, mammographic dental X-rays, well slices volumetric tomography resonance images....
In this paper, we propose a novel approach for few-shot semantic segmentation with sparse labeled images. We investigate the effectiveness of our method, which is based on Model-Agnostic Meta-Learning (MAML) algorithm, in medical scenario, where use labeling and can alleviate cost producing new annotated datasets. Our method uses labels meta-training dense meta-test, thus making model learn to predict from ones. conducted experiments four Chest X-Ray datasets evaluate two types annotations...