A5002110586- Anomaly Detection Techniques and Applications
- Mental Health via Writing
- Mental Health Research Topics
- Video Surveillance and Tracking Methods
- Stuttering Research and Treatment
- Digital Mental Health Interventions
- Functional Brain Connectivity Studies
- Hearing Impairment and Communication
- Fire Detection and Safety Systems
- Topic Modeling
- Text and Document Classification Technologies
- Vehicle License Plate Recognition
- Interpreting and Communication in Healthcare
- Schizophrenia research and treatment
- Language, Discourse, Communication Strategies
- Mental Health Treatment and Access
- Breastfeeding Practices and Influences
- Language Development and Disorders
- Network Security and Intrusion Detection
- Face and Expression Recognition
- Context-Aware Activity Recognition Systems
- Speech and dialogue systems
- Multimodal Machine Learning Applications
- Translation Studies and Practices
- Digital Communication and Language
Universidade Federal da Bahia
2021-2025
Universidade de São Paulo
2021-2023
National University of Ireland, Maynooth
2021
Universidade Federal do Paraná
2021
Recent developments in artificial intelligence technologies have come to a point where machine learning algorithms can infer mental status based on someone’s photos and texts posted social media. More than that, these are able predict, with reasonable degree of accuracy, future illness. They potentially represent an important advance health care for preventive early diagnosis initiatives, aiding professionals the follow-up prognosis their patients. However, issues call major caution use such...
Abstract The Nurturing Care Framework (NCF) calls for establishing a global monitoring and accountability systems early childhood development (ECD). Major gaps to build low‐cost large‐scale ECD at the local level remain. In this manuscript, we describe process of selecting nurturing care indicators municipal from existing routine information develop Brazilian Early Childhood Friendly Index (IMAPI). Three methodological steps developed through participatory decision‐making were followed....
Our study aimed to develop a machine learning ensemble distinguish "at-risk mental states for psychosis" (ARMS) subjects from control individuals the general population based on facial data extracted video-recordings.
Context: With the advancement of smart cities, University environment demands surveillance camera systems to increase their monitoring capabilities prevent malicious behaviors without prohibiting people's circulation. Problem: Universities have large parking lots with many vehicles and face daily security problems, such as robberies kidnappings, due lack cameras capable detecting suspicious behavior alerting personnel. Solution: Our approach enhances in university by developing a system...
Verbal communication has key information for mental health evaluation. Researchers have linked psychopathology phenomena to some of their counterparts in natural-language-processing (NLP). We study the characterization subtle impairments presented early stages psychosis, developing new analysis techniques and a comprehensive map associating NLP features with full range clinical presentation.
Despite significant advancements and pervasive use of vision-language models, a paucity studies has addressed their ethical implications. These models typically require extensive training data, often from hastily reviewed text image datasets, leading to highly imbalanced datasets concerns. Additionally, initially trained in English are frequently fine-tuned for other languages, such as the CLIP model, which can be expanded with more data enhance capabilities but add new biases. The CAPIVARA,...
Este artigo apresenta um estudo sobre a utilização de Ferramentas Ética para IA (AIETs) no levantamento considerações éticas modelos linguagem em português. As AIETs visam ajudar desenvolvedores, empresas, governos e outras partes interessadas estabelecer confiança, transparência responsabilidade com suas tecnologias. Busca-se este discutir brevemente se as podem auxiliar desenvolvedores pensarem eticamente seus modelos. Tal foi realizado partir entrevistas quatro utilizando Harms Modeling...
Spoken language is a key source of information for thought disorder evaluation. In the last decades, researchers linked psychopathology phenomena to their counterparts in natural processing (NLP) analysis. Nonetheless, seemingly opposite traits remain unconciliated. For instance, psychotic speech comprises incoherent trails, but also highly associated ones. order address some remaining gaps, we leveraged procedures from dynamical systems and graph theory. We examined transcribed interviews...
Context: With the growth of smart cities, surveillance camera systems have increased their monitoring capacity in different environments. In public Universities, requires subtle and precise alerts, order to curb actions that may present a lack security for academic community. Problem: Although detection objects videos is possible, high cost acquiring, installing high-quality cameras machines capable executing high-precision models customized solutions expensive. addition, most achieve...
The employment of video surveillance cameras by public safety agencies enables incident detection in monitored cities using object for scene description, enhancing the protection to general public. Object has its drawbacks, such as false positives. Our work aims enhance and image classification employing IoU (Intersection over Union) minimize positives identify weapon holders or fire a frame, adding more information scene.
Background and objective: To prevent the development of schizophrenia, preclinical stages disorder, known as "at-risk mental states for psychosis" (ARMS), have been intensively researched past three decades. Despite many advances in field, identification ARMS is still resource-consuming presents important issues regarding accuracy. address this, our study aimed to develop a machine learning ensemble distinguish from control individuals based on facial expression extracted brief...
Muitas técnicas têm sido desenvolvidas em prol da vigilância automatizada, desde alarmes inteligentes aos sistemas de câmeras. A automatização análise imagens vídeos, principalmente a detecção objetos tem crescido nos últimos tempos. Este trabalho por principal objetivo detectar específicos para auxiliar no alerta sobre porte armas, incêndios e pichações, tempo real, órgãos Secretaria Segurança Pública Bahia.
Network intrusion detection systems (NIDS) are one of many solutions that make up a computer security system. Several machine learning-based NIDS have been proposed in recent years, but most them were developed and evaluated under the assumption training context is similar to test context. In real networks, this false, given emergence new attacks variants known attacks. To deal with reality, open set recognition field, which general task recognizing classes not seen during any domain, began...
Natural language processing (NLP) techniques analyze formal properties of speechin humans. One way analyzing is through graph embeddings. Graphs are cunningly simple abstractions, which have vertices and nodes as basic elements. Nevertheless, graphs were revealed to be very useful abstractions describe associations in real-world phenomena whenever there coupled Graph embeddings associate words with adjacency a phrase links (edges). When repeated, the respective node linked twice. Some...
Abstract Background: Nonverbal communication (NVC) is a complex behavior that involves different modalities are impaired in schizophrenia spectrum, including gesticulation. However, there few studies evaluate it individuals with at-risk mental states (ARMS) for psychosis, mostly developed countries. Given our prior findings of reduced movement during speech seen Brazilian ARMS, we now aim to determine if this can be accounted by gesticulation behavior. Methods: 56 medication-naïve ARMS and...
Abstract Nonverbal communication (NVC) is a complex behavior that involves different modalities are impaired in the schizophrenia spectrum, including gesticulation. However, there few studies evaluate it individuals with at-risk mental states (ARMS) for psychosis, mostly developed countries. Given our prior findings of reduced movement during speech seen Brazilian ARMS, we now aim to determine if this can be accounted by gesticulation behavior. Fifty-six medication-naïve ARMS and 64 healthy...
<sec> <title>UNSTRUCTURED</title> Recent developments in artificial intelligence technologies have come up to a point where machine learning algorithms can infer mental status based on someone's photos and texts posted social media. More than that, these are able predict, with reasonable degree of accuracy, future illness. They potentially represent an important advance healthcare for preventive early diagnosis initiatives, aiding professionals the follow-up prognosing their patients....
Support Vector Machines are a set of machine learning models that have great performance in several tasks as well on image classification and object recognition. However, the proper choice model's hyperparameters has influence outcomes general capacity performance. In this paper, we explore some different traditional auto-tuning processes to estimate σ hyper-parameter for SVMs Gaussian kernel. These common also implemented standard software data science languages. The paper considers...