publications

2022

1. CAPTION: Caption Analysis with Proposed Terms, Image of Objects, and Natural Language Processing

   Leonardo Anjoletto Ferreira, Douglas De Rizzo Meneghetti, Paulo Eduardo Santos, and 1 more author

   SN Computer Science, Jul 2022

   Abs DOI

   This paper proposes a novel algorithm, called CAPTION, for identifying and correcting errors in automatically generated image captions. The algorithm combines Deep Learning (DL) for object detection in images with Natural Language Processing techniques. CAPTION has been tested in the following three tasks: (1) classify a caption as correct or not; (2) detect wrong words in the caption, and (3) suggest text corrections. Results show that our method is superior with respect to others evaluated in the same data set in the error correction task. These other methods are generally based exclusively on DL models. This work shows that, although semantics still has not been used at its fullest in this type of task, a combination of DL with Natural Language Processing tools presents a better overall performance than using DL methods alone.

2. Aprendizado Por Reforço Profundo Com Redes Recorrentes Aplicado à Negociação Do Minicontrato Futuro de Dólar

   Jonathan Kinoshita, Douglas Meneghetti, and Reinaldo Bianchi

   In Anais Do I Brazilian Workshop on Artificial Intelligence in Finance, Jul 2022

   DOI

2021

1. Detecting Soccer Balls with Reduced Neural Networks: A Comparison of Multiple Architectures under Constrained Hardware Scenarios

   Douglas De Rizzo Meneghetti, Thiago Pedro Donadon Homem, Jonas Henrique Renolfi de Oliveira, and 3 more authors

   Journal of Intelligent & Robotic Systems, Mar 2021

   Abs DOI

   Object detection techniques that achieve state-of-the-art detection accuracy employ convolutional neural networks, implemented to have optimal performance in graphics processing units. Some hardware systems, such as mobile robots, operate under constrained hardware situations, but still benefit from object detection capabilities. Multiple network models have been proposed, achieving comparable accuracy with reduced architectures and leaner operations. Motivated by the need to create an object detection system for a soccer team of mobile robots, this work provides a comparative study of recent proposals of neural networks targeted towards constrained hardware environments, in the specific task of soccer ball detection. We train multiple open implementations of MobileNetV2 and MobileNetV3 models with different underlying architectures, as well as YOLOv3, TinyYOLOv3, YOLOv4 and TinyYOLOv4 in an annotated image data set captured using a mobile robot. We then report their mean average precision on a test data set and their inference times in videos of different resolutions, under constrained and unconstrained hardware configurations. Results show that MobileNetV3 models have a good trade-off between mAP and inference time in constrained scenarios only, while MobileNetV2 with high width multipliers are appropriate for server-side inference. YOLO models in their official implementations are not suitable for inference in CPUs.

2. Specializing Inter-Agent Communication in Heterogeneous Multi-Agent Reinforcement Learning Using Agent Class Information

   Douglas De Rizzo Meneghetti, and Reinaldo Augusto da Costa Bianchi

   In AAAI-21 Workshop on Reinforcement Learning in Games, Feb 2021
3. Teste adaptativo informatizado da Provinha Brasil – Leitura: resultados e perspectivas

   Rodrigo Travitzki, Ocimar Munhoz Alavarse, Douglas De Rizzo Meneghetti, and 1 more author

   Estudos em Avaliação Educacional, Feb 2021

   Abs DOI

   Este artigo descreve um Teste Adaptativo Informatizado (TAI) para a Provinha Brasil – Leitura, com base na Teoria da Resposta ao Item. Detalha-se o funcionamento e processo de desenvolvimento do algoritmo. O TAI foi aplicado com o uso de tablets a 1.983 alunos dos 1o e 2o anos do ensino fundamental, em 15 escolas da Rede Municipal de Ensino de São Paulo. Os resultados confirmam a qualidade dos itens da Provinha Brasil, do trabalho realizado nas escolas e, sobretudo, do TAI. Em relaç ão à gestão do tempo de prova, conclui-se que há uma associação de mesmo sentido entre proficiência e tempo , mas só até certo ponto; os alunos tendem a demorar mais nos itens mais difíceis; essa tendência é mais intensa nos alunos mais proficientes, confirmando a hipótese de que eles tendem a gerir melhor o tempo de prova.

2020

1. Open Soccer Ball Dataset

   Reinaldo Augusto da Costa Bianchi, Danilo Hernani Perico, Thiago Pedro Donadon Homem, and 2 more authors

   Sep 2020

   Abs DOI

   Dataset with images of soccer ball acquired by a humanoid robot competing in the RoboCup Humanoid Kidsize League.

2. Annotated Image Dataset of Household Objects from the RoboFEI@ Home Team

   Douglas De Rizzo Meneghetti, Pedro Henrique Silva Domingues, Bruno de Freitas Vece Perez, and 6 more authors

   Sep 2020

   Abs DOI

   Annotated image dataset of household objects from the RoboFEI@Home teamThis data set contains two sets of pictures of household objects, created by the RoboFEI@Home team to develop object detection systems for a domestic robot.The first data set was created with objects from a local supermarket. Product brands are typical from Brazil. The second data set is composed of objects from the RoboCup@Home 2018 OPL competition.

3. Towards Heterogeneous Multi-agent Reinforcement Learning with Graph Neural Networks

   Douglas De Rizzo Meneghetti, and Reinaldo Augusto da Costa Bianchi

   Oct 2020
4. Towards Heterogeneous Multi-Agent Reinforcement Learning with Graph Neural Networks

   Douglas De Rizzo Meneghetti, and Reinaldo Augusto da Costa Bianchi

   In Anais Do Encontro Nacional de Inteligência Artificial e Computacional (ENIAC), Oct 2020

   Abs DOI

   This work proposes a neural network architecture that learns policies for multiple agent classes in a heterogeneous multi-agent reinforcement setting. The proposed network uses directed labeled graph representations for states, encodes feature vectors of different sizes for different entity classes, uses relational graph convolution layers to model different communication channels between entity types and learns distinct policies for different agent classes, sharing parameters wherever possible. Results have shown that specializing the communication channels between entity classes is a promising step to achieve higher performance in environments composed of heterogeneous entities.

2019

1. HERA: Home Environment Robot Assistant

   Plinio Thomaz Aquino Junior, Bruno de Freitas Vece Perez, Douglas De Rizzo Meneghetti, and 11 more authors

   In II Brazilian Humanoid Robot Workshop (BRAHUR) and III Brazilian Workshop on Service Robotics (BRASERO), Apr 2019
2. Object Detection under Constrained Hardware Scenarios: A Comparative Study of Reduced Convolutional Network Architectures

   Jonas Henrique Renolfi de Oliveira, Isaac Jesus da Silva, Thiago Pedro Donadon Homem, and 4 more authors

   In 2019 Latin American Robotics Symposium (LARS), 2019 Brazilian Symposium on Robotics (SBR) and 2019 Workshop on Robotics in Education (WRE), Oct 2019

   DOI
3. CAPTION: Correction by Analyses, POS -Tagging and Interpretation of Objects Using Only Nouns

   Leonardo Anjoletto Ferreira, Douglas De Rizzo Meneghetti, and Paulo Eduardo Santos

   In First Annual International Workshop on Interpretability: Methodologies and Algorithms (IMA 2019), Dec 2019

2018

1. Teste Adaptativo Informatizado Como Recurso Tecnológico Para Alfabetização Inicial

   Ocimar Munhoz Alavarse, Erica Maria de Toledo Catalani, Douglas De Rizzo Meneghetti, and 1 more author

   Revista Iberoamericana de Sistemas, Cibernética e Informá tica, Dec 2018
2. Comparação Entre Métodos de Detecção de Objetos Por Pontos-Chave e Redes Neurais

   Pedro Henrique Silva Domingues, Douglas De Rizzo Meneghetti, and Plinio Thomaz Aquino Junior

   In VIII Simpósio de Iniciação Científica, Didática e de Ações Sociais Da FEI, Dec 2018
3. Application and Simulation of Computerized Adaptive Tests Through the Package Catsim

   Douglas De Rizzo Meneghetti, and Plinio Thomaz Aquino Junior

   arXiv:1707.03012 [stat], Jul 2018

   Abs

   This paper presents catsim, the first package written in the Python language specialized in computerized adaptive tests and the logistical models of Item Response Theory. catsim provides functions for generating item and examinee parameters, simulating tests and plotting results, as well as enabling end users to create new procedures for proficiency initialization, item selection, proficiency estimation and test stopping criteria. The simulator keeps a record of the items selected for each examinee as well as their answers and also enables the simulation of linear tests, in which all examinees answer the same items. The various components made available by catsim can also be used in the creation of third-party testing applications. Examples of such usages are also presented in this paper.

2017

1. Teste Adaptativo Informatizado Como Recurso Tecnológico Para Avaliação Da Alfabetização Inicial

   Ocimar Munhoz Alavarse, Érica Maria de Toledo Catalani, Douglas Rizzo Meneghetti, and 1 more author

   In Séptima Conferencia Iberoamericana de Complejidad, Informá tica y Cibernética, Jul 2017

2016

1. RoboFEI@ Home 2016 Team Description Paper

   Andrey A Masiero, Douglas De Rizzo Meneghetti, Leonardo Contador, and 3 more authors

   Jul 2016

2015

1. Metodologia de Seleção de Itens Em Testes Adaptativos Informatizados Baseada Em Agrupamento Por Similaridade

   Douglas De Rizzo Meneghetti

   Jul 2015

2012

1. Desenvolvimento de Um Sistema de Aplicação de Testes Informatizados Com Conteúdo Multimídia

   Bruno de Jesus Destro, and Douglas De Rizzo Meneghetti

   Jul 2012