Startup linked to Unicamp is a pioneer in offering training in Brazil.
Next opportunity will be in São Paulo, starting on the 7th
In recent years, Artificial Intelligence has been gaining importance and generating ever greater expectations, starting from the unprecedented use in 2012, in a well-known Computer Vision competition called ImageNet, of an artificial intelligence technique called deep learning (DL). Deep learning is a disruptive machine learning technology with high performance in solving complex problems and flexibility in applying its algorithms. Among the main applications are image, voice and text recognition; predicting events and developing recommendation systems, for example to make the customer experience unique; and anomaly detection, for example for fraud detection.
O deep learning It presents enormous advantages when compared to classic machine learning techniques and even surpasses human performance in problems involving the analysis of large volumes of unstructured data. Giants like Google, Microsoft and Facebook are betting all their money on DL, and the forecast is that this market will be worth almost 2 billion dollars in five years. In Brazil, however, there are still very few companies that use DL.
With the aim of accelerating the adoption of deep learning In Brazil, Roberto Lotufo, retired full professor and collaborating researcher at the Faculty of Electrical and Computer Engineering (FEEC) at Unicamp, former director of the University's Innovation Agency and pioneer in offering extension courses in DL, created the company NeuralMind , in partnership with Patricia Magalhães de Toledo, a production engineer with extensive experience in innovation management, who also directed the Unicamp Innovation Agency. “This is an area that is revolutionizing several sectors. As deep learning, the computer can interpret an image better than a human being, it can achieve voice recognition and text processing on a previously unthinkable scale. In Brazil, however, there is still a lot of confusion between expert systems, machine learning and deep learning, with few people being able to distinguish the differences”, comments Lotufo.
In this context, NeuralMind brings together experts in artificial intelligence dedicated to developing and implementing solutions for different types of organizations, which involves, among other activities, offering courses. On the 7th and 21st of October and 4th of November, in São Paulo – at Chácara Santo Antônio, through a partnership with Spread Tecnologia and Genova Empreendedorismo –, the course “Deep Learning in Practice”, with a total workload of 24 hours and classes on Saturdays, from 9 am to 18 pm. The goals are to advance the understanding of the fundamentals of DL and design and train convolutional neural networks through the development of applications written in Python. As it is an Advanced level course, the target audience is made up of application developers with degrees in Engineering, Computer Science and Data Science. “We want participants to be able to develop real applications. I always use these two ingredients, theory and practice, together in my courses. Thus, theoretical concepts are assimilated through programming and carrying out practical experiments”, describes Lotufo.
Another difference is the qualification of the teaching staff. In addition to Lotufo – who is also a researcher at the Brazilian Institute of Neuroscience and Neurotechnology (BRAINN Fapesp) –, the team responsible for the content includes Rubens C. Machado, a retired researcher from the Renato Archer Information Technology Center with experience in neural networks since the 1980, and Rodrigo Nogueira, currently doing his doctorate at New York University, in the United States. In 2015, the three won the international liveness detection competition in fingerprint LivDet, with a solution using deep learning. In addition to them, Yugo Kuno, a doctoral student in Electrical Engineering at Unicamp, participates.
“I believe that we are the only ones to offer face-to-face and regular theoretical-practical courses in deep learning in Brazil. There are free online courses, but we understand that our in-person courses allow better monitoring and understanding of the participant, in addition to the practical opportunity to monitor the development of a project and training a deep neural network for real problems. We try to keep up with the state of the art in deep learning and offer the same quality as international courses in Brazil”, concludes Lotufo.
More information about the course, including program and registration procedures, is on the NeuralMind website, at neuralmindtech.com.br. There are also two other courses that will be offered in 2017: “Understanding DL”, Beginner level, lasting 8 hours, in Campinas, on October 14th; and “DL Accelerated”, Intermediate level, with 24 hours, also in Campinas, on 30/11 and 1/2 and 12. There will also be a new offer from “Deep Learning in Practice” in Campinas, in November and December. Check it out at neuralmindtech.com.br.
