Sergey Nikolenko explores fundamental computer vision problems, key synthetic data technologies, and future directions and applications
Synthesis AI, a pioneer in synthetic data technologies, today announced Springer has published the book Synthetic Data for Deep Learning written by Head of Artificial Intelligence (AI), Sergey Nikolenko. The book is available for purchase on Amazon and Springer.
Synthetic data refers to computer-generated images and simulations used to train computer vision models. Sergey Nikolenko is a computer scientist specializing in machine learning and the analysis of algorithms. In addition to his role at Synthesis AI, Nikolenko serves as the Head of the Artificial Intelligence Lab at the Steklov Mathematical Institute at St. Petersburg, Russia. His previous research includes works on cryptography, theoretical computer science, and algebra.
Synthetic Data for Deep Learning discusses fundamental computer vision problems, both low-level and high-level, synthetic environments and datasets for outdoor and urban scenes (i.e. autonomous driving), indoor scenes (i.e. indoor navigation), aerial navigation, and simulation environments for robotics. Additionally, it touches upon applications of synthetic data outside computer vision. Springer Publisher states, “This is the first book on synthetic data for deep learning, and its breadth of coverage may render this book as the default reference on synthetic data for years to come.”
Serge Belongie, Professor, Department of Computer Science at the University of Copenhagen (DIKU) and Director, Pioneer Centre for Artificial Intelligence, said, “As deep learning finds its way into a rapidly growing array of real-world applications ranging from autonomous vehicles to telemedicine, the need for data to train ever-higher capacity models shows no signs of stopping. To meet that need, our field must tap into rich sources of synthetic and real data. Sergey’s book lucidly surveys the state of the art in the former, and I consider it required reading for any researcher using deep learning based methods.”
“The sheer pace — not to mention complexity — of machine learning and deep learning development is exponential in the world around us. I felt strongly that we needed a comprehensive text to look at the rise of synthetic data needed to prolong the exponential growth of machine learning in supervised learning problems, especially computer vision, and more,” said Nikolenko. “Simply put, synthetic data is a way to prolong the march of progress in these fields. I hope this text can help educate, serve as a comprehensive reference for all aspects of synthetic data, and facilitate discussion and future research.”
Coming out of stealth with $4.5M in funding in April 2021, Synthesis AI’s platform addresses industry needs by letting customers programmatically create vast amounts of perfectly labeled, unbiased image data enabling the development of more capable models. The company has since achieved the largest synthetic data set in the industry with 40K unique identities and delivered 10M labeled images from its FaceAPI product. Synthesis AI’s customers include top handset manufacturers, global technology companies, teleconferencing companies, and leading chipset and camera manufacturers. The company also recently announced enhanced capabilities to enable the development of driver safety monitoring systems and is already working with leading automobile manufacturers.
For more such updates and perspectives around Digital Innovation, IoT, Data Infrastructure, AI & Cybsercurity, go to AI-Techpark.com.