- Smart Agriculture and AI
- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Machine Learning and Algorithms
- Spectroscopy and Chemometric Analyses
- Machine Learning and Data Classification
- Genomics and Phylogenetic Studies
- Video Analysis and Summarization
- RNA and protein synthesis mechanisms
- Soybean genetics and cultivation
- Advanced Image Fusion Techniques
- AI in cancer detection
- Algorithms and Data Compression
- Visual Attention and Saliency Detection
- Chromosomal and Genetic Variations
- Advanced Neural Network Applications
- Anomaly Detection Techniques and Applications
- Gene expression and cancer classification
- Advanced Graph Neural Networks
- Imbalanced Data Classification Techniques
- Multimodal Machine Learning Applications
- Leaf Properties and Growth Measurement
- Video Surveillance and Tracking Methods
- Cutaneous Melanoma Detection and Management
- Agricultural and Food Sciences
Universidade Federal de São Carlos
2022-2024
Universidade Tecnológica Federal do Paraná
2014-2023
Universidade Estadual de Campinas (UNICAMP)
2013-2022
Hospital Universitário da Universidade de São Paulo
2020
Centro Universitário Eurípedes de Marilia
2020
Vigor (United States)
2018
Secretaria Municipal de Cultura
2018
Universidade de São Paulo
2009
Brazilian Society of Computational and Applied Mathematics
2009
National Council for Scientific and Technological Development
2009
Abstract Transposable elements (TEs) are the most represented sequences occurring in eukaryotic genomes. Few methods provide classification of these into deeper levels, such as superfamily level, which could useful and detailed information about sequences. Most that classify TE use handcrafted features k-mers homology-based search, be inefficient for classifying non-homologous Here we propose an approach, called transposable pepresentation learner (TERL), preprocesses transforms...
Due to datasets have continuously grown, efforts been performed in the attempt solve problem related large amount of unlabeled data disproportion scarcity labeled data. Another important issue is trade-off between difficulty obtaining annotations provided by a specialist and need for significant annotated obtain robust classifier. In this context, active learning techniques jointly with semi-supervised are interesting. A smaller number more informative samples previously selected (by...
Nowadays, there is an abundance of biomedical data, such as images and genetic sequences, among others. However, a lack annotation to volume due the high costs involved perform this task. Thus, it mandatory develop techniques ease burden human annotation. To reach goal active learning strategies can be applied. state-of-the-art methods, generally, are not feasible lead with real-world datasets. Another important issue, that generally neglected by these related conception classifier tends...
Nowadays, large datasets are common and demand faster more effective pattern analysis techniques. However, methodologies to compare classifiers usually do not take into account the learning-time constraints required by applications. This work presents a methodology with respect their ability learn from classification errors on learning set, within given time limit. Faster techniques may acquire training samples, but only when they will achieve higher performance unseen testing sets. We...
Agribusiness has a great relevance in the world׳s economy. It generates considerable impact gross national product of several nations. Hence, it is major driver many economies. Nowadays, from each new planting to harvesting process mandatory and crucial apply some kind technology optimize given singular process, or even entire cropping chain. For instance, digital image analysis joined with machine learning methods can be applied obtain guarantee higher quality harvest, leading not only...
Nowadays, deep features, obtained from a variety of learning architectures, play an important role in several real problems. It is know that transfer strategies could be employed to take advantage such features trained under general context (e.g. ImageNet). However, the best our knowledge, majority works focus on similar contexts accomplish strategies. Thus, this work we analyze content-based medical image retrieval, and demonstrate it possible make use specific context, like mammographic...
Very high resolution (VHR) images are large datasets for pixel annotation -- a process that has depended on the supervised training of an effective classifier. Active learning techniques have mitigated this problem, but descriptors limited to local image information and number pixels makes response time user's actions impractical, during active learning. To circumvent we present strategy relies superpixel priori dataset reduction. Firstly, compare VHR using superpixel- pixel-based...
The labor-intensive and time-consuming process of annotating data is a serious bottleneck in many pattern recognition applications when handling massive datasets. Active learning strategies have been sought to reduce the cost on human annotation, by means automatically selecting most informative unlabeled samples for annotation. critical issue lies selection such samples. As an effective solution, we propose active approach that preprocesses dataset, efficiently reduces organizes set selects...
Seed companies increasingly seek excellence in production quality through rigorous processes, such as the tetrazolium test (TZ test) and vigor definition. However, these are extremely laborious processes since it necessitates experience of a specialist visual analysis considerable quantity seeds sampling for determining seed lot.Moreover, although TZ has defined protocol, this may vary from analyst to because is subjective human process. In context, several efforts have been carried out an...
Abstract Transposable elements (TEs) are the most represented sequences occurring in eukaryotic genomes. They capable of transpose and generate multiple copies themselves throughout These can produce a variety effects on organisms, such as regulation gene expression. There several types these elements, which classified hierarchical way into classes, subclasses, orders superfamilies. Few methods provide classification deeper levels, superfamily level, could useful detailed information about...
The main drawback regarding agriculture commodities such as soybeans is that they must be uniform in quality considering the companies produce and sell it. Thus, to reach requirements it necessary establish strict parameters control processes. most important test accomplish this task based on seed vigor definition. However, great majority of analysis performed by a human specialist leading an extremely tiresome subjective approach, highly susceptible failures, well as, certain types...
Recently, impressive results have been provided by pre-trained convolutional neural networks combined with the transfer learning technique. They quickly become a great option to classify general image datasets. However, best of our knowledge, majority works do not explore if these architectures are well-suited specific contexts like medical domain (e.g. breast lesions). We focus on lesions, because it is one most common types cancer affecting women worldwide, and its early diagnosis crucial...
One of the main issues related to treatment soybean seeds concerns its vigor definition. The great majority analysis is performed by a human specialist leading an extremely tiresome and subjective approach, highly susceptible failures. In order deal with this problem present paper proposed methodology joining content-based image retrieval techniques perception retrieve according his/her expectation aiding seed experiments showed that approach presented several notable contributions...
Soy is the main product of Brazilian agriculture and fourth most cultivated bean globally. Since soy cultivation tends to increase due this large market, guarantee quality an indispensable factor for enterprises stay competitive. Industries perform vigor tests acquire information evaluate planting. The tetrazolium test, example, provides about moisture damage, bedbugs, or mechanical damage. However, verification damage reason its severity are done by analyst, one one. massive exhausting...