Leannet : uma arquitetura que utiliza o contexto da cena para melhorar o reconhecimento de objetos

• Ano de defesa: 2018
• Autor(a) principal: Silva, Leandro Pereira da
• Orientador(a): Ruiz, Duncan Dubugras Alcoba
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica do Rio Grande do Sul
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação
• Departamento: Escola Politécnica
• País: Brasil
• Palavras-chave em Português: Detecção de Objetos; Rede Neural Convolucional; Rede Neural; Aprendizagem Profunda; Objetos em Contexto
• Palavras-chave em Inglês: Object Detection; Convolutional Neural Network; Neural Network; Deep Learning; Object in Context
• Área do conhecimento CNPq: CIENCIA DA COMPUTACAO::TEORIA DA COMPUTACAO
• Link de acesso: http://tede2.pucrs.br/tede2/handle/tede/8168

Computer vision is the science that aims to give computers the capability of see- ing the world around them. Among its tasks, object recognition intends to classify objects and to identify where each object is in a given image. As objects tend to occur in particular environments, their contextual association can be useful to improve the object recognition task. To address the contextual awareness on object recognition task, the proposed ap- proach performs the identification of the scene context separately from the identification of the object, fusing both information in order to improve the object detection. In order to do so, we propose a novel architecture composed of two convolutional neural networks running in parallel: one for object identification and the other to the identification of the context where the object is located. Finally, the information of the two-streams architecture is concatenated to perform the object classification. The evaluation is performed using PASCAL VOC 2007 and MS COCO public datasets, by comparing the performance of our proposed approach with architectures that do not use the scene context to perform the classification of the ob- jects. Results show that our approach is able to raise in-context object scores, and reduces out-of-context objects scores.