Scientists have developed a microchip that can capture visual details from video frames using extremely low levels of power. According to the research team, the chip's video feature extractor uses 20 times less power than existing best-in-class chips. The new microchip, called EQSCALE, could reduce the size of smart vision systems down to the millimeter range. Current chips consume up to hundreds of milliwatts to power a smartphone, for example. To enable near-perpetual operation, devices can be powered by solar cells that harvest energy from natural lighting in living spaces. However, such devices would require solar cells with a size in the centimeter scale or larger, limiting miniaturization. For systems to be scaled down to the millimeter scale, they would need to consume less than one milliwatt of power, according to the team. A team of researchers led by associate professor Massimo Alioto from the Department of Electrical and Computer Engineering at the NUS Faculty of Engineering has developed a tiny vision processing chip, EQSCALE, which uses 20 times less power than existing technology. Courtesy of National University of Singapore. The National University of Singapore (NUS) team's microchip can perform continuous feature extraction at 0.2 mW. This could translate into an advancement in the level of miniaturization that is currently possible for smart vision systems. The new feature extractor is smaller than a millimeter on each side. “This technological breakthrough is achieved through the concept of energy-quality scaling, where the trade-off between energy consumption and quality in the extraction of features is adjusted," NUS associate professor Massimo Alioto said. "This mimics the dynamic change in the level of attention with which humans observe the visual scene, processing it with different levels of detail and quality depending on the task at hand. Energy-quality scaling allows correct object recognition even when a substantial number of points of interests are missed due to the degraded quality of the target.” EQSCALE, which can be powered continuously by a millimeter-sized solar cell without the need for battery replacement, could be a step toward millimeter-sized vision systems that could operate indefinitely. The team is investigating the possibility of developing a miniaturized computer vision system that comprises smart cameras equipped with vision capabilities enabled by the microchip and a machine learning engine that comprehends the visual scene. The ultimate goal of the team is to enable massively distributed vision systems for wide-area and ubiquitous visual monitoring, vastly exceeding the traditional concept of cameras. “Vision electronic systems with long lifetimes are currently not feasible for IoT applications due to their high power consumption and large size," said Alioto. "Our team has addressed these challenges through our tiny EQSCALE chip, and we have shown that ubiquitous and always-on smart cameras are viable. We hope that this new capability will accelerate the ambitious endeavor of embedding the sense of sight in the IoT,” The research was published in IEEE Xplore (doi: 10.1109/ASSCC.2017.8240261) and presented at the 2017 IEEE Asian Solid-State Circuits Conference, Nov. 6-8, 2017 in Seoul, South Korea.