PhD CIFRE, PhD – Deep-based reference pictures generation for video codin

About InterDigital InterDigital is a global research and development company focused primarily on wireless, video, artificial intelligence (“AI”), and related technologies. We design and develop foundational technologies that enable connected, immersive experiences in a broad range of communications and entertainment products and services. We license our innovations worldwide to companies providing such products and services, including makers of wireless communications devices, consumer electronics, IoT devices, cars and other motor vehicles, and providers of cloud-based services such as video streaming. As a leader in wireless technology, our engineers have designed and developed a wide range of innovations that are used in wireless products and networks, from the earliest digital cellular systems to 5G and today’s most advanced Wi-Fi technologies. We are also a leader in video processing and video encoding/decoding technology, with a significant AI research effort that intersects with both wireless and video technologies. Founded in 1972, InterDigital is listed on Nasdaq. InterDigital is a registered trademark of InterDigital, Inc. For more information, visit:  www.interdigital.com . InterDigital R&I Center in Rennes, France, is looking for a PhD researcher in deep-based video coding. The offer takes place in a research project focused on 2D video coding, deeply involved in standardization (MPEG, VCEG, JVET). The InterDigital team has a very strong presence and recognition in MPEG and JVET in the conventional and deep-learning based video coding fields and activities. If deep-learning solutions are starting to demonstrate high-performance coding efficiency, they still remain below the state-of-the-art conventional solutions (for example, the latest MPEG video coding standard named VVC, or even more the currently exploratory model developed by JVET, named ECM). The objective of the PhD thesis is to explore new coding algorithms based on deep-learning solutions, focused on the improvement of interprediction by using deep solutions to generate more accurate reference pictures. The main aim is to significantly improve the coding efficiency compared to the prior-art solutions. The challenge is to develop performing deep-learning based coding tools, that are generic enough so that they can be inserted into a conventional video coding architecture, as well as in a full end-to-end deep-based video coding solution. Complexity issues are also key in this work. The PhD researcher will develop and implement innovative deep-based video coding algorithms, using reference software implementations of the most recent technologies developed by JVET or MPEG. The researcher will have opportunities to participate to the external promotion of R&I technologies via contribution to standards, publications, conferences. A high scientific understanding is required to execute her/his role in a team of InterDigital experts and in the external scientific…