Search

Vision chips perform image capture and real-time intelligent image processing by integrating an imager and a vision processor on a single chip, having broad application prospects. This brief proposes a programmable and flexible vision processor with a dual-issue micro-architecture. The processor consists of a reconfigurable vector unit, a flexible memory access network, and a non-maximum suppression (NMS) block. It can efficiently implement both deep neural network (DNN) and traditional computer vision (CV) algorithms. The vector unit performs single-instruction multiple-vector (SIMV) parallel operations with reconfigurable vector width. The flexible memory access network adaptively supports multiple vector operations under different vector widths. A four-MAC processing element (PE) in the vector unit is designed to increase computational power and data reuse rate. The NMS block can speed up the object location processing of the detection networks. The chip is fabricated in a 28nm process. The experimental results show that the maximum clock frequency, peak performance, and peak energy efficiency are 600MHz, 1.2TOPS, and 2.03TOPS/W, respectively. The Mobilenet-V1 processing achieves a throughput of 404 fps under a $256\times 224$ image size and an 87.15%(top-5) accuracy on the ImageNet dataset. [ABSTRACT FROM AUTHOR]

Nowadays, the vision chip bridging sensing and processing has been extensively employed in high-speed image processing, owing to its excellent performance, low power consumption, and economical cost. However, there is a dilemma in designing processors to support conventional computer vision algorithms and neural networks since the two algorithms have a non-trivial trade-off in proposing a unified architecture. By analyzing computation properties, we propose a novel hierarchical parallel vision processor (ViP) for hybrid vision chips to accelerate both traditional computer vision (CV) and neural network (NN). The ViP architecture includes three parallelism levels: PE for pixel-centric, computing core (CC) for block, and vision core (VC) for global. PEs contain dedicated computing units and data paths for convolution operations without degrading its flexibility. Each CC is driven by customized SIMD instructions and can be dynamically connected for meeting block parallelism requirements. ViP is fabricated in 65nm CMOS technology and achieves a peak performance of 614.4 GOPS and an energy efficiency of 640 GOPS/W at 200 MHz clock frequency. Notably, several experiments on CV and NN are performed, illustrating an ultra-low latency in executing hybrid algorithms. [ABSTRACT FROM AUTHOR]

All-CMOS monolithic microdisplay technologies have been attracting attention due to their direct integration of light-emitting pixel arrays and driving circuits on a single silicon substrate. Improvements to optical power efficiency have been a hot spot for all-silicon microdisplay technologies. MOS-like gate-control structure avalanche-mode light-emitting diodes (AMLEDs) that employ hot-carrier electroluminescence to produce light emission are used to enhance the optical power efficiency of pixel units. A semi-active matrix mode is used to reduce the power consumption of driver circuits. A 100 × 100 pixel array with an all-CMOS monolithic microdisplay system is demonstrated using standard 0.18 μm CMOS technology. The optical emission power and the breakdown voltage of the proposed AMLEDs are increased by 139.2% and reduced by more than 67%, respectively. The optical-power efficiency is 5.98 × 10−7 which is comparable to the best reported so far in all-CMOS monolithic microdisplay chip. With its significant optical-power efficiency, the proposed microdisplay technology has broad application prospects in near-to-eye display and head mounted displays (HMDs). [ABSTRACT FROM AUTHOR]

With the growing number of research studies on autonomous driving, a lot of new technologies for self-driving cars are also constantly emerging. While significantly promoting the development of autonomous vehicles, these emerging technologies also have potential vulnerabilities, which may bring up some nonnegligible security threats. Loopholes existing in localization and navigation technologies, which are the key elements of route planning and navigation, could be utilized by adversaries to manipulate autonomous driving navigation by hijacking valuable vehicles, goods, or even target characters. Toward this end, we provide in this article a basic guide summarizing the security threats and attack schemes, and we propose countermeasures specifically for localization and navigation technologies in autonomous vehicles. After that, we present a novel route spoofing attack that could also be implemented against future autonomous driving navigation. With such implementation, we demonstrate that effectively, routes of targeted drivers can be usurped and hence mislead drivers to take a wrong route, go by a malicious place, and even be tracked in real time. [ABSTRACT FROM AUTHOR]

The rapid development of IoT, cloud computing, Artificial Intelligence (AI), big data, and 5G technologies has promoted the transformation of traditional vehicles toward ICVs. Compared to traditional vehicles, ICVs face many security threats introduced by network technologies and intelligent devices, especially in the field of remote wireless communications using Telematics Service Provider (TSP). As the core communication system in ICVs, TSP integrates diverse communication systems, and thus inherits the original vulnerabilities of these systems inevitably. TSP provides various methods for the ICVs to access the Internet, which makes them vulnerable to remote attacks. However, existing auto manufacturers mostly focused on the user experiences of the ICVs, and paid little attention to these potential security risks raised by TSP. Toward this end, in this article we analyze and summarize the TSP security threats in ICVs, and present some attack methodologies. After that, we discuss a practical attack case against an ICV by leveraging the vulnerabilities of TSP, and some countermeasures are proposed to enhance ICV security against TSP attacks. [ABSTRACT FROM AUTHOR]

Intelligent and Connected Vehicles (ICVs), which are equipped with various wireless communication technologies and many smart systems, significantly improve the driving experience. Nevertheless, due to the potential vulnerabilities in their advanced cyber physical features and growing numbers of interfaces, malicious attacks can be easily implemented toward ICVs, in particular by utilizing the interfaces of wireless telematics systems. Remote vehicle attacks can even be launched by a malicious application or a smart connected device. Toward this end, we provide in this article an essential tutorial summarizing the threats, attacking methodologies, and promising solutions specifically for wireless telematics systems in the emerging ICVs. [ABSTRACT FROM AUTHOR]