A5006191190- Helicobacter pylori-related gastroenterology studies
- Gastric Cancer Management and Outcomes
- Cholangiocarcinoma and Gallbladder Cancer Studies
- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Anomaly Detection Techniques and Applications
- Fuzzy Logic and Control Systems
- Medical Image Segmentation Techniques
- Gene expression and cancer classification
- Neural Networks and Applications
- Network Security and Intrusion Detection
- Reproductive Biology and Fertility
- Fuzzy Systems and Optimization
- Advanced Vision and Imaging
- Internet Traffic Analysis and Secure E-voting
- Remote Sensing and LiDAR Applications
- Handwritten Text Recognition Techniques
- Image and Object Detection Techniques
- Assisted Reproductive Technology and Twin Pregnancy
- Face and Expression Recognition
- Natural Language Processing Techniques
- Topic Modeling
- Sparse and Compressive Sensing Techniques
- Video Surveillance and Tracking Methods
- Advanced Malware Detection Techniques
Universidade de São Paulo
2015-2025
Akebono Clinic
2004-2024
Institute of Mathematics and Computer Science
2023
Czech Academy of Sciences, Institute of Mathematics
2022
Instituto Butantan
2021
Institute for Systems Engineering and Computers
2021
University of Coimbra
2021
Instituto de Telecomunicações
2021
University of Aveiro
2021
University of Pittsburgh
2021
We present a comprehensive study and evaluation of existing single image deraining algorithms, using new large-scale benchmark consisting both synthetic real-world rainy images.This dataset highlights diverse data sources contents, is divided into three subsets (rain streak, rain drop, mist), each serving different training or purposes. further provide rich variety criteria for dehazing algorithm evaluation, ranging from full-reference metrics, to no-reference subjective the novel...
Abstract Lung cancer is the leading cause of deaths in United States, surpassing breast as primary cancer-related mortality women. The goal present study was to identify early molecular changes lung induced by exposure tobacco smoke and thus potential targets for chemoprevention. Female A/J mice were exposed either or HEPA-filtered air via a whole-body chamber (6 h/d, 5 d/wk 3, 8, 20 weeks). Gene expression profiles tissue from control smoke-exposed animals established using 15K cDNA...
The Future Internet will integrate large-scale systems constructed from the composition of thousands distributed services, while interacting directly with physical world via sensors and actuators, which compose Things. This enable realization Smart Cities vision, in urban infrastructure be used to its fullest extent offer a better quality life for citizens. Key efficient effective is scientific technological research covering multiple layers that make up Internet. paper discusses challenges...
Abstract High incidence of gastric cancer-related death is mainly due to diagnosis at an advanced stage in addition the lack adequate neoadjuvant therapy. Hence, new tools aimed early would have a positive impact outcome disease. Using cDNA arrays having 376 genes either identified previously as altered tumors or known be human cancer, we determined expression signature 99 tissue fragments representing normal mucosa, gastritis, intestinal metaplasia, and adenocarcinomas. We first validated...
Security is a critical issue in the context of IoT and, more recently, Industrial IoT(IIoT) environments. To mitigate security threats, Intrusion Detection Systems have been proposed. Still, most them can achieve high accuracy only by having access to application layer flows, which problematic terms privacy. This paper presents neural network model based on hybrid CNN-LSTM architecture detect several attacks traffic at Edge IIoT using features from transport and layers. Besides improving...
Abstract Adenocarcinomas of stomach and esophagus are frequently associated with preceding inflammatory alterations the normal mucosa. Whereas intestinal metaplasia gastric mucosa is higher risk malignization, Barrett's disease a factor for adenocarcinoma esophagus. characterized by substitution squamous columnar tissue classified histopathologically as metaplasia. Using cDNA microarrays, we determined expression profile esophageal well adenocarcinomas from both organs. Data were explored to...
Persistent infection by high risk HPV types (e.g. HPV-16, -18, -31, and -45) is the main factor for development of cervical intraepithelial neoplasia cancer. Tumor necrosis (TNF) a key mediator epithelial cell inflammatory response exerts potent cytostatic effect on normal or HPV16, but not HPV18 immortalized keratinocytes. Moreover, several carcinoma-derived lines are resistant to TNF anti-proliferative suggesting that acquisition TNF-resistance may constitute an important step in...
Visualizing decision boundaries of machine learning classifiers can help in classifier design, testing and fine-tuning. Decision maps are visualization techniques that overcome the key sparsity-related limitation scatterplots for this task. To increase trustworthiness map use, we perform an extensive evaluation considering dimensionality-reduction (DR) projection underlying construction. We extend visual accuracy by proposing additional to suppress errors caused distortions. Additionally,...
Cyber-attacks and threats are growing fast in the Internet of Things (IoT) infrastructure as applications smart cities gain momentum. Usually, IoT devices communicate via machine-to-machine protocols such Message Queuing Telemetry Transport (MQTT). Due to heterogeneous structure absence security by design methodologies, mechanisms environments with MQTT traffic needed, they can be deployed Intrusion Detection Systems (IDS). This paper proposes a Deep Learning (DL) based Network IDS trained...
An important aspect of mathematical morphology is the description complete lattice operators by a formal language, Morphological Language (ML), whose vocabulary composed infimum, supremum, dilations, erosions, anti-dilations and anti-ero
Vehicle detection in video is an important problem Computer Vision because of the potential applications security, vehicle traffic, driving assistance and so on. In this work, we used Mixture Deformable Part Models (MDPM) for sequences obtained from static dynamic cameras. The MDPM method was originally proposed by Felzenszwalb et al realm object images. We tested detection. designed a set experiments that explore number components mixture parts model. performed comparison study symmetric...
This article introduces a method for road network extraction from satellite images. The proposed approach covers new fusion (using data multiple sources) and Markov random field (MRF) defined on connected components along with multilevel application (two-level MRF). Our allows the detection of roads different characteristics decreases by around 30% size used graph model. Results synthetic aperture radar (SAR) images optical obtained using TerraSAR-X Quickbird sensors, respectively, are...
For convolutional neural networks (CNNs), a common hypothesis that explains both their generalization capability and characteristic brittleness is these models are implicitly regularized to rely on imperceptible high-frequency patterns, more than humans would do. This has seen some empirical validation, but most works do not rigorously divide the image frequency spectrum. We present model spectrum in disjointed discs based distribution of energy apply simple feature importance procedures...
Abstract BACKGROUND Nodules of the thyroid gland are observed frequently in patients who undergo ultrasound studies. The majority these nodules benign, corresponding to goiters or adenomas, and only a small fraction corresponds carcinomas. Among tumors, diagnosis follicular adenocarcinomas by preoperative fine‐needle aspiration biopsy is major challenge, because it requires inspection entire capsule differentiate from adenoma. Consequently, large numbers unnecessary thyroidectomy. METHODS...
Abstract Estrogen acts via its receptor (ER) to stimulate cell growth and differentiation in the mammary gland. ER progesterone (PR), which is regulated by estrogen ER, have been used as prognostic markers clinical management of breast cancer patients. Patients with − tumors a poorer prognosis than patients + tumors. The aim present study was identification tumor‐associated genes differentially expressed regarding presence or absence PR hybridized cDNA microarrays containing 4,500...
Staff line removal is an important pre-processing step to convert content of music score images machine readable formats. Many heuristic algorithms have been proposed for staff and recently a competition was organized in the 2013 ICDAR/GREC conference. Music are often subject different deformations variations, existing do not work well all cases. We investigate application learning based method problem. The consists multiple image operators from training input-output pairs then combining...