The Defense Advanced Research Projects Agency (DARPA) has released a Special Notice (SN) to get more information about researching and developing “third wave” AI theory, and ways to address limitations seen in first and second wave AI technologies.
DARPA’s AI Exploration (AIE) program, which the agency said has been a “key element of DARPA’s broader AI investment strategy that will help ensure the U.S. maintains a technological advantage” in the AI space, will be leading the charge.
According to the SN, “AIE will enable DARPA to fund pioneering AI research to discover new areas where R&D programs awarded through this new approach may be able to advance the state of the art,” and “will enable DARPA to go from idea inception to exploration in 90 days.”
The SN details that the DARPA Microsystems Technology Office (MTO) is interested in a potential AIE program topic dealing with In Pixel Intelligent Processing (IP2). IP2 technology will help with the accuracy and functionality of deep neural networks (NNs) “in power-constrained sensing platforms, with 10x fewer operations compared to state-of-the-art (SOA) NNs.”
“AI processing of video presents a challenging problem because high resolution, high dynamic range, and high frame rates generate significantly more data in real time than other edge sensing modalities,” the SN states. “The number of parameters and memory requirement for SOA AI algorithms typically is proportional to the input dimensionality and scales exponentially with the accuracy requirement,” it says.
The SN adds that “IP2 will seek to solve two key elements required to embed AI at the sensor edge: data complexity and implementation of accurate, low-latency, low size, weight, and power (SWaP) AI algorithms.”
“IP2 will require performers to demonstrate SOA accuracy with 20x reduction of AI algorithm processing energy delay product while processing complex datasets” like the University of California-Berkley’s BDD100K – a self-driving car dataset that demonstrates third wave functionality by incorporating geographic, environmental and weather diversity, intentional occlusions and a large number of classification tasks.
