César Roberto De Souza | Naver Labs Europe
loader image
Change your cover photo
Change your cover photo
blank
This user account status is Approved

This user has not added any information to their profile yet.

Computer Vision

César has obtained his cum laude PhD from the Universitat Autonoma de Barcelona, researching methods for human action recognition in videos. He also holds a Master's degree on Computer Science specialized in Image and Signal Processing from the Universidade Federal de São Carlos, after researching and creating methods for the automatic recognition of signs and words from the Brazilian Sign Language in videos.

 


 

Before becoming a computer vision researcher, César had accumulated deep expertise on .NET and related Microsoft technologies with 10+ years experience in software development, with knowledge on the internal workings of the managed runtime, unsafe code, handling of unmanaged resources and interoperability with unmanaged code. As both a researcher and a software developer, he had created the Accord.NET Framework, a framework for scientific computing, mathematics, statistics, computer vision, computer audition, machine learning, and artificial intelligence. The framework has been featured in three books, has been cited in a large number of scientific publications, and was used daily by professionals, practitioners, and hobbyists alike.

However, César now dedicates his time entirely to research, as a member of Computer Vision group @ NLE.

  • Jerome Revaud, Jon Almazan, Rafael Sampaio de Rezende, and Cesar Roberto de Souza. 2019. “Learning with Average Precision: Training Image Retrieval with a Listwise Loss.” In International Conference on Computer Vision 2019 (ICCV’19).
  • Jerome Revaud, Philippe Weinzaepfel, César De Souza, Noe Pion, Gabriela Csurka, Yohann Cabon, and Martin Humenberger. 2019. “R2D2: Repeatable and Reliable Detector and Descriptor.” ArXiv Preprint ArXiv:1906.06195.
  • César Roberto de Souza, Adrien Gaidon, Yohann Cabon, Naila Murray, and Antonio Manuel López. 2019. “Generating Human Action Videos by Coupling 3D Game Engines and Probabilistic Graphical Models.” International Journal of Computer Vision, 1–32.
  • Jerome Revaud, César Roberto de Souza, Martin Humenberger, and Philippe Weinzaepfel. 2019. “R2D2: Reliable and Repeatable Detector and Descriptor.” In Advances in Neural Information Processing Systems, 12405–15.
  • César Roberto de Souza, Adrien Gaidon, Eleonora Vig, and Antonio M Lopez. 2018. “System and Method for Video Classification Using a Hybrid Unsupervised and Supervised Multi-Layer Architecture.”
  • César Roberto de Souza. 2018. “Action Recognition in Videos: Data-Efficient Approaches for Supervised Learning of Human Action Classification Models for Video.” PhD Thesis, Barcelona.
  • César Roberto de Souza, Boris Chidlovskii, Frederic Roulland, and Luis Rafael Ulloa Paredes. 2018. “Methods and Systems for Interpretable User Behavior Profiling in Off-Street Parking.”
  • César Roberto de Souza, Luis Rafael Ulloa Paredes, Boris Chidlovskii, and Victor Ciriza. 2017. “Method, System and Processor-Readable Media for Estimating Airport Usage Demand.”
  • Luis Rafael Ulloa Paredes, and César Roberto de Souza. 2017. “System and Method for Generating Available Ride-Share Paths in a Transportation Network.”
  • César Roberto de Souza, Frederic Roulland, and Victor Ciriza. 2017. “Method and System for Simulating Users in the Context of a Parking Lot Based on the Automatic Learning of a User Choice Decision Function from Historical Data Considering Multiple User Behavior Profiles.”
  • César Roberto de Souza, Adrien Gaidon, Yohann Cabon, and Antonio Manoel Lopez Pena. 2017. “Procedural Generation of Videos to Train Deep Action Recognition Networks.” In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
  • César Roberto de Souza, Adrien Gaidon, Eleonora Vig, and Antonio Manuel López. 2016. “Sympathy for the Details: Dense Trajectories and Hybrid Classification Architectures for Action Recognition.” In European Conference on Computer Vision, 697–716.
  • Cesar Roberto de Souza, Boris Chidlovskii, Frederic Roulland, and Luis Rafael Ulloa Paredes. 2016. “Methods and Systems for Interpretable User Behavior Profiling in Off-Street Parking.”
  • Renata Mendes, César Roberto de Souza, Maurício Machado, Paulo Correa, Luciana Di Thommazo-Luporini, Ross Arena, Jonathan Myers, Ednaldo Brigante Pizzolato, and Audrey Borghi-Silva. 2015. “Predicting Reintubation, Prolonged Mechanical Ventilation and Death in Post-Coronary Artery Bypass Graft Surgery: A Comparison between Artificial Neural Networks and Logistic Regression Models.” Archives of Medical Science: AMS 11 (4): 756.
  • Jonathan Myers, Cesar Roberto de Souza, Audrey Borghi-Silva, Marco Guazzi, Paul Chase, Daniel Bensimhon, Mary Ann Peberdy, et al. 2014. “A Neural Network Approach to Predicting Outcomes in Heart Failure Using Cardiopulmonary Exercise Testing.” International Journal of Cardiology 171 (2): 265–69.
  • Frederic Roulland, Cesar Roberto de Souza, Luis Ulloa, Arturo Mondragon, Michael Niemaz, and Victor Ciriza. 2014. “Towards Data-Driven Simulations in Urban Mobility Analytics.” In 14th ITS Asia-Pacific Forum 2015 Nanjing, China.
  • César Roberto de Souza. 2014. “The Accord.NET Framework.” http://accord-framework.net.
  • César Roberto de Souza, and Ednaldo Brigante Pizzolato. 2013. “Sign Language Recognition with Support Vector Machines and Hidden Conditional Random Fields: Going from Fingerspelling to Natural Articulated Words.” In International Workshop on Machine Learning and Data Mining in Pattern Recognition, 84–98.
  • Ross Arena, Jonathan Myers, Cesar Roberto de Souza, Audrey Borghi Silva, Marco Guazzi, Paul Chase, Daniel Bensimhon, et al. 2013. “A Neural Network Approach to Predicting Outcomes in Heart Failure Using Cardiopulmonary Exercise Testing.” Journal of the American College of Cardiology 61 (10 Supplement): E1415.
  • César Roberto de Souza 2013. “Reconhecimento de Gestos Da Língua Brasileira de Sinais Através de Máquinas de Vetores de Suporte e Campos Aleatórios Condicionais Ocultos.” Masters thesis. Universidade Federal de São Carlos, São Carlos.
  • César Roberto de Souza, Ednaldo Pizzolato, and Mauro dos Santos Anjo. 2012. “Fingerspelling Recognition with Support Vector Machines and Hidden Conditional Random Fields.” Advances in Artificial Intelligence--IBERAMIA 2012, 561–70.
  • César Roberto de Souza, Ednaldo Brigante Pizzolato, and Mauro dos Santos Anjo. 2012. “Recognizing Static Signs from the Brazilian Sign Language: Comparing Large-Margin Decision Directed Acyclic Graphs, Voting Support Vector Machines and Artificial Neural Networks.” ArXiv Preprint ArXiv:1210.7461.
  • César Roberto de Souza. 2012. “A Tutorial on Principal Component Analysis with the Accord.NET Framework.” ArXiv Preprint ArXiv:1210.7463.
  • César Roberto de Souza. 2012. “Automatic Image Stitching with Accord.NET.”
  • Sebastian Feuerstack, Jessica Colnago, César Roberto de Souza, Ednaldo Brigante Pizzolato, and others. 2011. “Designing and Executing Multimodal Interfaces for the Web Based on State Chart XML.” In Proceedings of Third W3C Web Conference Brasil.
  • César Roberto de Souza. 2010. “Kernel Functions for Machine Learning Applications.” http://crsouza.org.
  • César Roberto de Souza. 2010. “Kernel Support Vector Machines for Classification and Regression in C#.” http://crsouza.org.
  • Cesar Roberto de Souza, Ednaldo Pizzolato, Renata Mendes, Audrey Borghi-Silva, Mauríicio Machado, and Paulo Correa. 2009. “Artificial Neural Networks Prognostic Evaluation of Post-Surgery Complications in Patients Underwent to Coronary Artery Bypass Graft Surgery.” In 2009 International Conference on Machine Learning and Applications, 799–803.
  • Cesar Roberto de Souza, Ednaldo Brigante Pizzolatto, Audrey Borghi-Silva, Renata Mendes, Maurício Machado, and Paulo Correa. 2009. “Avaliação Prognóstica de Complicações Pós-Operatórias Por Meio de Redes Neurais Em Pacientes Submetidos à Cirurgia de Revascularização Do Miocárdio.” In IX Congresso Brasileiro de Redes Neurais / Inteligência Computacional (IX CBRN).
  • Cesar Roberto de Souza. 2009. “Neural Network Learning by the Levenberg-Marquardt Algorithm with Bayesian Regularization.” http://crsouza.org.

César is a certified web tutor by the Federal University of São Carlos, having taught in:

  • e-course on Information Storage and Retrieval, Universidade Federal de São Carlos (UFSCar), 2013
  • e-course on Object Orientated Programming, Universidade Federal de São Carlos (UFSCar), 2013
  • e-course on Information Organization and Retrieval, Universidade Federal de São Carlos (UFSCar), 2013
  • e-course on Information Organization and Retrieval, Universidade Federal de São Carlos (UFSCar), 2014
  • e-course on Information Storage and Retrieval, Universidade Federal de São Carlos (UFSCar), 2015
0476614141