The Internet-of-Things (IoT) is expected to connect everyone and everything (video, audio, texting, and so on) in a seamless network. It will lead to unprecedented explosion of data. Among these data, the video data is increasingly becoming the “biggest big data.” The current systems will suffer from huge cost in transfer and store the huge volume of data. Hence, the efficiency of video data processing mechanism is going to be challenged. 1) We need to employ highly efficient data compression systems for transmitting and keeping the enormous volumes of data. 2) Since not all of the data are useful, we will face huge challenges to improve data learning/understanding capabilities to distinguish between useful and redundant data. 3) Energy efficiency will be essential to save the power consumption of the loT devices.

　Intelligent Media Processing Laboratory focuses on the algorithms and VLSI architectures for multimedia signal processing. 1) We deal with measurement coding algorithms for compressively sensed visual contents to realize a batteryless video acquisition system. 2) We develop the techniques to realize low-power high-performance VLSI design for video coding. 3) We do research on the machine learning technologies for learning and understanding the video data.