Close
Metadata

@InProceedings{BatistadaCunhaSanValMagTei:2020:StImDo,
               author = "Batista da Cunha, Kelvin and dos Santos Brito, Caio Jos{\'e} and 
                         Valen{\c{c}}a da Rocha Martins Albuquerque, Lucas and 
                         Magalh{\~a}es Sim{\~o}es, Francisco Paulo and Teichrieb, 
                         Veronica",
          affiliation = "Voxar Labs, Centro de Inform{\'a}tica, Universidade Federal de 
                         Pernambuco and Voxar Labs, Centro de Inform{\'a}tica, 
                         Universidade Federal de Pernambuco and Voxar Labs, Centro de 
                         Inform{\'a}tica, Universidade Federal de Pernambuco and Curso 
                         T{\'e}cnico em Inform{\'a}tica para Internet, Instituto Federal 
                         de Pernambuco, Campus Belo Jardim and Voxar Labs, Centro de 
                         Inform{\'a}tica, Universidade Federal de Pernambuco",
                title = "A Study on the Impact of Domain Randomization for Monocular Deep 
                         6DoF Pose Estimation",
            booktitle = "Proceedings...",
                 year = "2020",
               editor = "Musse, Soraia Raupp and Cesar Junior, Roberto Marcondes and 
                         Pelechano, Nuria and Wang, Zhangyang (Atlas)",
         organization = "Conference on Graphics, Patterns and Images, 33. (SIBGRAPI)",
            publisher = "IEEE Computer Society",
              address = "Los Alamitos",
             keywords = "Pose Estimation, Deep Learning, Domain Randomization.",
             abstract = "In this work, we apply domain randomization to synthetic images 
                         and train deep 6DoF monocular RGB pose estimation models to work 
                         on a real object. We compare 19 models trained with different 
                         combinations of synthetic and real data (fully synthetic, fully 
                         real, initially synthetic and supplemented with real, and a 
                         real-synthetic randomized mix). By gradually decreasing the amount 
                         of real data used, we show it is possible for deep 6DoF detection 
                         to obtain superior results while using less real data (which is 
                         harder to obtain) and suggest the best approach to train a model 
                         with synthetic data. Our method is validated using a textureless 
                         3D printed object, as the textureless category is a challenging, 
                         common open problem in itself. A real and a synthetic dataset 
                         generated for this work, totalling over 24,800 annotated frames, 
                         are also made public. We also show that synthetic, randomized data 
                         can help generalize a model by training it to handle challenges 
                         such as illumination changes and fast motion. Finally, we also 
                         evaluate how a model trained for one camera sensor works with a 
                         different one, and show that synthetic simulations of real cameras 
                         can help overcoming this challenge.",
  conference-location = "Virtual",
      conference-year = "Nov. 7-10, 2020",
             language = "en",
           targetfile = "108.pdf",
        urlaccessdate = "2020, Dec. 04"
}


Close