The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
The method of Graph Cuts converts a Maximum a Posteriori (MAP) inference problem on a Markov Random Field (MRF) into a network flow, which can be solved efficiently. Many computer vision problems can be conveniently cast as an inference task to find most likely labels for pixels. The method is widely used, but computationally burdensome. Prior accelerator attempts have failed to exploit the problem's...
The growing prominence and computational challenges imposed by Deep Neural Networks (DNNs) has fueled the design of specialized accelerator architectures and associated dataflows to improve their implementation efficiency. Each of these solutions serve as a datapoint on the throughput vs. energy trade-offs for a given DNN and a set of architectural constraints. In this paper, we set out to explore...
Deep Neural Networks (DNNs) have emerged as a powerful and versatile set of techniques showing successes on challenging artificial intelligence (AI) problems. Applications in domains such as image/video processing, autonomous cars, natural language processing, speech synthesis and recognition, genomics and many others have embraced deep learning as the foundation. DNNs achieve superior accuracy for...
A design of flexible piezoelectric strain energy harvester responsive to multi-directional forces from arbitrary human motions was developed by using polydimethylsiloxane (PDMS) and polyvinylidene fluoride (PVDF). Unlike the most of conventional strain energy harvesters designed to be functional only for single directional motion, our suggested design demonstrated the energy harvesting capability...
This paper firstly reports a highly reliable MEMS-switch employing a CNTs-network lubricant in the contact-area. By covering the contact-area with the CNTs-network, we achieved more than an order of magnitude extension in the lifetime of the MEMS-switch. We determined that this drastic improvement in the reliability arises from the compressibility of the CNTs-network, generating remarkable contact-area...
We developed a novel approach to fabricate suspended metal- and metal oxide nanowires using carbon nanotubes (CNTs) as sacrificial template. As a potential application that harnesses the unique properties of suspended nanowires, sensitive chemical sensors were demonstrated. The suspended CNTs template was synthesized on the side wall of the electrodes by chemical vapor deposition (CVD), followed by...
We present an energy-efficient implementation of RGB-D simultaneous localization and mapping (SLAM) by applying approximate computing (AC) techniques such as loop perforation (LP) and reduced precision (RP). To reduce processing time and power consumption, LP and RP were applied to the two most computationally challenging portion of the multi-kernel pipeline of SLAM, i.e., SIFT feature detection/extraction...
Approximate computing is gaining traction as a computing paradigm for data analytics and cognitive applications that aim to extract deep insight from vast quantities of data. In this paper, we demonstrate that multiple approximation techniques can be applied to applications in these domains and can be further combined together to compound their benefits. In assessing the potential of approximation...
We demonstrate a novel FPGA-based accelerator architecture that can tackle a range of standard computer vision (CV) problems, with scalable performance and attractive speedups. The architecture relies on multiple pipelined processing elements (PEs) that can be configured to support various belief propagation (BP) settings for different CV tasks. Inside each PE, innovative implementation of Jump Flooding...
Probabilistic inference is a versatile tool to solve a large variety of pixel-labeling problems in computer vision such as stereo matching and image denoising. Belief Propagation (BP) is an effective method for such inference tasks, and has also shown attractive error-resilience properties—the ability to converge to usable solutions in the presence of low-level hardware errors. This is of increasing...
Maximum a posteriori probability (MAP) inference on Markov random fields (MRF) is the basis of many computer vision applications. Sequential tree-reweighted belief propagation (TRW-S) has been shown to provide very good inference quality and strong convergence properties. However, software TRW-S solvers are slow due to the algorithm's high computational requirements. A state-of-the-art FPGA implementation...
In this paper, we present an error resilient Markov random field (MRF) message passing based stereo matching hardware (HW) architecture. Previously, algorithmic noise tolerance (ANT) has been applied at the arithmetic level of the reparameterize unit and showed greatly enhanced robustness of message passing inference based architectures. In this work, correction was targeted at the system level to...
Message passing based inference algorithms have immense importance in real-world applications. In this paper, error resilience of a message passing based Markov random field (MRF) stereo matching architecture is explored and enhanced through application of algorithmic noise tolerance (ANT) in order to cope with nanometer imperfections in post-silicon devices. We first explore the inherent robustness...
In this paper, we present our hardware accelerator for inference computations on Markov Random Fields (MRFs), which wins the “adjusted run time” category of MEMOCODE 2013 design contest. The contest problem is to accelerate the popular Belief Propagation (BP) algorithm for MRF stereo matching, but BP often suffers from non-convergence in its MRF inference. To overcome the drawbacks of BP, we show...
We have developed a highly sensitive, transparent and flexible toluene sensor using cobalt-metalloporphyrin (Co-MPP)-functionalized graphene as sensing material. The functionalization of MPP enhanced responsiveness (ΔR/Ri) over 200% compared to pristine graphene toward 10 ppm of toluene molecules at room temperature, while it still revealed optical transmittance and mechanical flexibility.
In this paper, we describe hardware for inference computations on Markov Random Fields (MRFs). MRFs are widely used in applications like computer vision, but conventional software solvers are slow. Belief Propagation (BP) solvers, which use patterns of local message passing on MRFs, have been studied in hardware, but their performance is unreliable. We show how a superior method—Sequential Tree-Reweighted...
We have demonstrated length-controllable in-plane synthesis of aligned carbon nanotube (CNT) array on microfabricated structures using micromechanical springs. The micromechanical spring provides precise compressive stress during the chemical vapor deposition process for CNT growth. Different loading results in different final length of the CNT array, as well as different alignment and defectiveness...
We demonstrated a novel usage of self-adjusted, vertically aligned carbon nanotube (CNT) arrays integrated on the sidewalls of microstructures as latching components. The CNT array-based latching mechanism showed stable latching at multiple latching positions, together with reversible and bidirectional latching capabilities. The latchable shuttle using CNT latch could be adopted for diverse microelectromechanical...
A reliable electromechanical memory device using aligned carbon nanotube (CNT) arrays as contact material has been developed. The aligned CNT arrays were integrated by chemical vapor deposition (CVD) on microstructures making mechanical contact. The contact showed hysteretic pull-out and — in behaviors which enabled on — and off-state readout, and the memory logic was programmed based on adhesive...
A microswitch based on self-assembled carbon nanotube (CNT) arrays as micromechanical contact material has been demonstrated. The aligned CNT arrays are synthesized on microelectrodes and movable shuttle fabricated on a silicon-on-insulator (SOI) wafer. The CNT arrays are self-assembled on microstructures making mechanical contact between source and shuttle, and this contact is preloaded by the growth...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.