A5064860556- Face recognition and analysis
- Smart Agriculture and AI
- Speech and Audio Processing
- Remote Sensing in Agriculture
- Remote Sensing and LiDAR Applications
- 3D Shape Modeling and Analysis
- Assistive Technology in Communication and Mobility
- Seismic Imaging and Inversion Techniques
- Online Learning and Analytics
- Advanced Image and Video Retrieval Techniques
- Disability Education and Employment
- Impact of Technology on Adolescents
- Plant and animal studies
- Child Development and Digital Technology
- Digital Accessibility for Disabilities
- Seismology and Earthquake Studies
- Education Pedagogy and Practices
- Telemedicine and Telehealth Implementation
- Education and Public Policy
- Tactile and Sensory Interactions
- Seismic Waves and Analysis
- Surgical Simulation and Training
- Digital Imaging in Medicine
- Image Retrieval and Classification Techniques
- Entrepreneurship Studies and Influences
Microsoft (United States)
2022-2023
IBM Research - Brazil
2014-2020
IBM (United States)
2014
Universidade de São Paulo
2010
The population growth and consequent global rise in food demand require increasingly efficient agricultural solutions, what is commonly called digital agriculture. Among promising initiatives, the use of remotely sensed data combined with machine learning algorithms enables handling faster operations lower associated cost. One most important activities agriculture crop identification, which fundamental for managing inventory a farm by producers governmental authorities, has been addressed...
BackgroundThe COVID pandemic brought the need for more realistic remote consultations into focus. 2D Telemedicine solutions fail to replicate fluency or authenticity of in-person consultations. This research reports on an international collaboration participatory development and first validated clinical use a novel, real-time 360-degree 3D system worldwide. The - leveraging Microsoft's Holoportation™ communication technology – commenced at Canniesburn Plastic Surgery Unit, Glasgow, in March...
Much work has been done on the assessment of texture descriptors for image retrieval in many domains. In this work, we evaluate accuracy and performance three wellknown - Gabor Filters, GLCM, LBP seismic retrieval. These subsurface images pose challenges yet not thoroughly investigated previous works, which are addressed evaluated our experiments. We asked domain experts to annotate two cubes, Penobscot 3D Netherlands F3, used them descriptors, corresponding parameters, similarity metrics...
In this paper, we report on a qualitative study that investigates the impact of using popular Massive Open Online Course (MOOC) to complement vocational training students with intellectual disability (ID). We have been investigating problem for several months in partnership Brazilian NGO (Non-Governmental Organization) people ID. Our methodology integrates different aspects human-computer interaction (i.e., requirement gathering sessions and observation real subjects). Potential users were...
Vocational training of people with disabilities (PwD) can potentially improve social and economic prospects, but at the same time, it be significantly challenging due to need for specialized technology. Unfortunately, in developing countries this problem is magnified because, general, low-income groups have limited access appropriate content assistive technologies. In paper, we present initial findings from a qualitative field study computer-mediated vocational students intellectual...
In this paper, we report on our experiences investigating the role of digital technology in face-to-face instruction students with intellectual disability. process, used a multi-method approach and findings integrate results from focus groups, interviews, observations, iterative prototyping, user evaluation. Ultimately, hope that work can motivate future research efforts bring to light opportunities be considered development mobile-based education solutions.
Deep learning methods have become the standard for Visual Speech Recognition problems due to their high accuracy results reported in literature. However, while successful works been words and sentences, recognizing shorter segments of speech, like phones, has proven be much more challenging lack temporal contextual information. Also, head-pose variation remains a known issue facial analysis with direct impact this problem. In context, we propose novel methodology tackle problem visemes –...
Previous work demonstrated that people who rely on lip-reading often prefer a frontal view of their interlocutor, but sometimes profile may display certain lip gestures more noticeably. This refers to an assistive tool receives unconstrained video speaker, captured at arbitrary view, and not only locates the mouth region also displays augmented versions lips in views. is made using deep Generative Adversarial Networks (GANs) trained several pairs images. In training set, each pair contains...
In this paper, we consider the use of technological instructional pacing supports to teach students with intellectual disability (ID). Based on a qualitative field study where 11 participants used our mobile-based educational platform, found that although technology may help instructor control pace class, it also poses barriers development students' autonomy and self-esteem. Our preliminary results suggest balance between instructor-led self-paced instruction based is promising would better...
Detecting salt domes in seismic images is very important for the exploration of petroleum reservoirs. However, this task characterized as being time-consuming when performed by human interpreters due to structural complexity bodies found on different volumes. This work aims at performing a comprehensive evaluation texture descriptors broadly used image processing community applied images. A robust multi-scale analysis conducted order assess which features and corresponding parameters are...
In this paper we introduce a Facial Animation system using real three-dimensional models of people, acquired by 3D scanner. We consider dataset composed displaying different facial expressions and linear interpolation technique is used to produce smooth transition between them. One-to-one correspondences the meshes each expression are required in order apply process. Instead focusing computation dense correspondence, some points selected triangulation defined, being refined consecutive...
Seismic interpretation is a complex procedure that depends on many and interdependent data analyses. One of the essential steps in this process picking horizons seismic images, which time-consuming prone to errors when performed manually. In context, having reliable horizon tool fundamental for accurate interpretation. Although several methods have been proposed literature tools made available industry, most require numerous iterations manual corrections delivering satisfactory results....
Vocational training can bring significant benefits for people with disabilities (PwD), particularly in terms of self-esteem and autonomy. Nevertheless, only a small fraction Brazilians actually work, due to lack job qualifications. In this paper, we report on the early progress an ongoing research agenda that investigates new educational social engagement technologies facilitate qualification inclusion PwD Brazilian labor market. Based our experiences working multiple disability populations...
In this work we present an image processing-based assistant for helping visually impaired citizens with the task of recognizing dynamic content within fixed layouts displays in public spaces. Our solution relies on placement markers, order to facilitate location and recognition target objects and, at same time, provide hints users about how better position their mobile device's cameras capture whole information contained display.
Digital education has potential to provide different possibilities for personalization and consequently reach a larger more diverse number of people. Personalization is key component solutions addressing important long-standing pedagogical challenges in education, such as dealing with heterogeneity learning styles. In particular scenarios where accessibility support required, depends on the creation representations individual pieces content. this light, main goal article describe how we...
Esse artigo descreve resultados preliminares de um estudo qualitativo conduzido com homens e mulheres empregados em uma empresa naárea tecnologia. Foram conduzidos quatro grupos focais dois objetivos principais: (i) identificar as influências mais relevantes que pessoas selecionadas receberam no momento optaram por curso superior tecnologia; (ii) verificar se esses fatores diferem entre profissionais do sexo masculino feminino. Nossas observações indicam o processo da escolha carreira difere...
The use of unmanned aerial vehicles (UAVs) and computer vision for automating farm operations is growing rapidly: time-consuming tasks such as crop monitoring may be solved in a more efficient, precise, less error-prone manner. In particular, estimating productivity managing pests, it fundamental to characterize regions into four classes: (i) full-grown trees, (ii) tree seedlings, (iii) gaps, (iv) background. this paper, we address the classification images from citrus plantations, acquired by...
This work describes an efficient approach for flower classification that is suitable deployment in mobile devices, allowing its use a citizen science application biodiversity monitoring. In the proposed system, geo-located images are uploaded by user and segmented semi-automatically. We propose method based on histogram comparison of color, shape texture cues, using metric learning feature weighting. Our tested Oxford Flower Dataset we able to achieve state-of-the-art accuracy, while...
The access to information displayed in public spaces is a challenge faced by visually impaired people for which image processing techniques have the potential deliver satisfactory solutions. However, object recognition algorithms must initially locate possible candidates images, hard task complex scenes. In this article, we introduce an technique that relies on incorporation of markers panels and boards with fixed layouts displaying dynamic content. allow: a) locating objects be recognized;...
Fez-se uma análise da concepção de Desenvolvimento Profissional Docente (DPD) que aparece nas Diretrizes Curriculares Nacionais, para a Formação dos Professores Educação Básica. O objetivo foi compreender do DPD presente no contexto das Resoluções CNE/CP n.º 2/2015, 2/2019 e 1/ 2020. A metodologia se fundamenta na epistemologia dialético-materialista enfoque qualitativo investigação educacional. No plano operacional, usaram-se pesquisa documental bibliográfica. primeira, exploração...
Visual Speech Recognition is the ability to interpret spoken text using video information only. To address such task automatically, recent works have employed Deep Learning and obtained high accuracy on recognition of words sentences uttered in controlled environments, with limited head-pose variation. However, drops for multi-view datasets when it comes interpreting isolated mouth shapes, as visemes, values reported are considerably lower, shorter segments speech lack temporal contextual...