Search for: All records

Computer vision on low-power edge devices enables applications including search-and-rescue and security. State-of-the-art computer vision algorithms, such as Deep Neural Networks (DNNs), are too large for inference on low-power edge devices. To improve efficiency, some existing approaches parallelize DNN inference across multiple edge devices. How-ever, these techniques introduce significant communication and synchronization overheads or are unable to balance workloads across devices. This paper demonstrates that the hierarchical DNN architecture is well suited for parallel processing on multiple edge devices. We design a novel method that creates a parallel inference pipeline for computer vision problems that use hierarchical DNNs. The method balances loads across the collaborating devices and reduces communication costs to facilitate the processing of multiple video frames simultaneously with higher throughput. Our experiments consider a representative computer vision problem where image recognition is performed on each video frame, running on multiple Raspberry Pi 4Bs. With four collaborating low-power edge devices, our approach achieves 3.21× higher throughput, 68% less energy consumption per device per frame, and a 58% decrease in memory when compared with existing sinaledevice hierarchical DNNs. 

Autonomous vehicles (AVs) use diverse sensors to understand their surroundings as they continually make safety-critical decisions. However, establishing trust with other AVs is a key prerequisite because safety-critical decisions cannot be made based on data shared from untrusted sources. Existing protocols require an infrastructure network connection and a third-party root of trust to establish a secure channel, which are not always available.In this paper, we propose a sensor-fusion approach for mobile trust establishment, which combines GPS and visual data. The combined data forms evidence that one vehicle is nearby another, which is a strong indication that it is not a remote adversary hence trustworthy. Our preliminary experiments show that our sensor-fusion approach achieves above 80% successful pairing of two legitimate vehicles observing the same object with 5 meters of error. Based on these preliminary results, we anticipate that a refined approach can support fuzzy trust establishment, enabling better collaboration between nearby AVs. 

As we add more autonomous and semi-autonomous vehicles (AVs) to our roads, their effects on passenger and pedestrian safety are becoming more important. Despite extensive testing before deployment, AV systems are not perfect at identifying hazards in the roadway. Although a particular AV’s sensors and software may not be 100% accurate at identifying hazards, there is an untapped pool of information held by other AVs in the vicinity that could be used to quickly and accurately identify roadway hazards before they present a safety threat. 

Low-power computer vision on embedded devices has many applications. This paper describes a low-power technique for the object re-identification (reID) problem: matching a query image against a gallery of previously-seen images. State-of-the-art techniques rely on large, computationally-intensive Deep Neural Networks (DNNs). We propose a novel hierarchical DNN architecture that uses attribute labels in the training dataset to perform efficient object reID. At each node in the hierarchy, a small DNN identifies a different attribute of the query image. The small DNN at each leaf node is specialized to re-identify a subset of the gallery---only the images with the attributes identified along the path from the root to a leaf. Thus, a query image is re-identified accurately after processing with a few small DNNs. We compare our method with state-of-the-art object reID techniques. With a ~4% loss in accuracy, our approach realizes significant resource savings: 74% less memory, 72% fewer operations, and 67% lower query latency, yielding 65% less energy consumption.