Wed, 25 May, 05:00 - 06:00 France Time (UTC +2)
Wed, 25 May, 03:00 - 04:00 UTC
Tue, 24 May, 23:00 - 00:00 New York Time (UTC -4)
Chair: Woon-Seng Gan, Nanyang Technological University, Singapore
Demand for improved customer experience drives the need for increasing levels of human-centric signal processing in consumer electronics. Humans care deeply about the quality of images, audio, and video – high quality capture and rendering are immensely important for addressing this care-about. Although our profession has pursued better quality images, audio, and video for a long time, there continues to be opportunity for improvement. New sensor, speaker, and display technologies give new capabilities, but simultaneously require more advanced algorithms and system solutions. Typically, this means the computational requirements also increase substantially. In this talk, we consider the current state of the art for capture and rendering. Specifically, we consider examples of current smartphone cameras, speakers, and displays. We look at the embedded signal processing currently applied including new AI-based solutions in the signal chain. Then, considering the challenges still remaining in these devices, we discuss what is needed for the next generation. The processing requirements in these devices must be balanced with cost, power, and other practical concerns. To design successful future products, we must carefully predict the computations required and create a next-generation design with precisely the amount of computations needed, without excessive performance and cost. To do this, we consider industry trends and the range of available processors that can be leveraged in consumer devices. By taking advantage of increasing computational performances of traditional processor cores, and additionally tapping into the exponentially increasing capabilities of the new compute engines such as neural processing units, we plan to continue the mission to further improve customer experience.
Dr. Mike Polley is Senior Vice President and Head of the Mobile Processor Innovation Lab at Samsung where he leads a team of world-class algorithm and system designers focused on creating advanced technologies for Samsung’s Galaxy smartphones as well as next-generation mobile devices.
Prior to Samsung, Mike worked at Texas Instruments for 18 years defining chipset architectures and leading embedded signal processing R&D. He was recognized for his technical accomplishments by election to TI Fellow in 2008.
Mike received his B.S., M.S., and Ph.D. degrees in electrical engineering from MIT. He holds 41 U.S. patents on a broad range of products across communications and multimedia systems.