Li Huang, Junjie Li, Weiwei Jiang, Zhiyu Zhang, Minchuan Chen, Shaojun Wang, Jing Xiao. Proceedings of the 59th Annual Meeting Association for Computational Linguistics and 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

This paper presents a clockless digital filter able to process inputs of different rates and formats, synchronous or asynchronous, with no adjustment needed handle each input type. The is designed using mix asynchronous real-time hardware, for this reason relies on neither clock nor the data rate setting its frequency response. modular architecture filter, including delay segments separated timing paths pipelined multi-way adder, allows easy extensions widths. was used as part an ADC/DSP/DAC...

An asynchronous high-performance low-power 5-port network-on-chip (NoC) router is introduced. The proposed integrates low-latency input buffers using a circular FIFO design, and novel end-to-end credit-based virtual channel (VC) flow control for replicated switch architecture. This then compared to an AMD synchronous router, in realistic advanced 14nm FinFET library. the first such comparison, best of our knowledge, real baseline already fabricated several commercial products. Initial...

The wine industry has a significant socioeconomic impact. Recent advancements in smart agriculture have greatly influenced vine growth and wine-making processes. In this study, we evaluate ratings by relating them to multiple characteristics. A novel dataset with various features is curated, an interpretable learning paradigm proposed. Several different models are employed, leading promising experimental results. Furthermore, through feature contribution analysis, discover that...

Infrared and visible image fusion can combine the thermal radiation information textures to provide a high-quality fused image. In this letter, we propose hybrid variational model achieve end. Specifically, an ℓ0 term is adopted preserve highlighted targets with salient gradient variation in infrared image, ℓ1 used suppress noise ℓ2 employed keep of Experimental results demonstrate superiority proposed our have more sharpen less noise.

The increasing demand for broadband wireless service in today's communications, especially with WiMAX technology being one of the 3G standards, is causing cellular network providers to consider integration different heterogeneous networks. Consequently, selection techniques play a vital role ensuring quality In this article we develop scheme an integrated UMTS/WiMAX system. design goal provide user best available QoS at any time. proposed integrates Variance Coefficient Weighting (VCW) and...

A new asynchronous low-latency interconnection network is introduced for a 2D mesh topology. The network-on-chip, named AEoLiAN, contains fast lightweight monitoring to notify the routers of incoming traffic, thereby allowing arbitration and channel allocation be initiated in advance. In contrast, several recent synchronous early methods require significant resource overhead, including use hybrid networks, or wide channels additional VCs. proposed approach has much smaller finer-grain...

In this article, a novel interconnect technology is presented for the cost-effective and flexible design of asynchronous networks-on-chip. It delivers asynchrony in heterogeneous system integration while yielding low-energy on-chip data movement. The approach consists both lightweight switch architecture (using transition-signaling protocols bundled-data encoding) complete synthesis flow built on top mainstream industrial CAD tools. For first time, article demonstrates compelling area,...

This paper presents a clockless digital filter able to process inputs of different rates and formats, synchronous or asynchronous, with no adjustment needed handle each input type. The 16-tap, 8-bit FIR filter, integrated in 130 nm CMOS process, includes on-chip automatic delay tuning. chip was used as part ADC/DSP/DAC chain which, unlike the case conventional, clocked systems, maintains its frequency response intact when sample rate changes. For certain inputs, system has signal-to-error...

Loan default prediction has been a significant problem in the financial domain because overdue loans may incur losses. Machine learning methods have introduced to solve this problem, but there are still many challenges including feature multicollinearity, imbalanced labels, and small data sample problems. To replicate success of deep areas, an effective regularization technique named muddling label is letter, ensemble feed-forward neural networks proposed, which outperforms machine baselines...

Since the dark channel prior (DCP)-based dehazing method is ineffective in sky area and will cause problem of too color distortion image, we propose a novel based on segmentation image fusion. We first segment according to characteristics non-sky then estimate atmospheric light transmission map DCP correct them, fuse original after contrast adaptive histogram equalization improve details information image. Experiments illustrate that our performs well can reduce distortion.

In this study, total alkaloids from Hemsleya chinensis were extracted and tested for their antioxidant properties. To optimize extraction methods, a single factor experiment was conducted to determine the alkaloid concentrations of H. using L9 (34) orthogonal design test method BP neural network (BPNN), resulting in optimum conditions alkaloids. The optimal with acid water are: HCl concentration is 0.50 %, temperature 85 °C, material-liquid ratio 1:64.5, rate 0.2785 ± 0.0003 mg/mL. exhibited...

There is increasing interest in using m-of-n delay-insensitive codes for robust asynchronous global communication, to support the design of coding-efficient and low-power channels. However, a fundamental obstacle these has been complex expensive hardware support. This paper addresses this issue, introducing evaluating practical completion detector units 2-of-n codes. Designs are proposed both return-to-zero (RZ) non-return-to-zero (NRZ) The RZ designs build on prior work Piestrak [14];...

A differential pulse code modulation (DPCM)-based system is proposed for lossless compression of medical images. Three techniques are used in the system. The intensity current pixel predicted optimally by linear combination intensities N pixels a neighborhood causal region. multilayer perceptron, feedforward neural network, to model nonlinearity hierarchical prediction scheme adapted progressive transmission. Coefficients each layer chosen achieve minimum entropy. Prediction errors, assumed...

Multicast communication (1-to-many) is common in parallel architectures and emerging areas such as neuromorphic computing. However, there very limited research supporting multicast asynchronous NoCs. This paper proposes a new NoC with 2D-mesh topology. To the best of our knowledge, this first general-purpose to support 2D meshes. A critical feature use continuous-time replication strategy, where flits packet are routed through distinct outputs router according each output's own rate,...

A new asynchronous arbitration node is introduced for use as a building block in an interconnection network. The target network topology variant Mesh-of-Trees (MoT), combining binary fan-out (i.e. routing) and fan-in arbitration) network, which becoming widely used multi-core shared-memory interfaces. two key features are: (i) each can resolve its pre-allocate the corresponding input channel, before actual data arrives; (ii) lightweight shadow monitoring fast forwards information soon enters...

Drought has been a global concern and an effective prediction method is needed. Meteorological data are seen as efficient economic approach. Challenges arise with the large volume high nonlinearity between meteorological variables drought level. In this study, deep learning proposed solution for level multivariate time series classification. The synthetic minority oversampling technique further adopted to alleviate class imbalance problem improve classification performance. Experimental...

This paper introduces the theory of knowledge supply chain, and places demand enterprises into whole through identifying gaps enterprise management, at different levels, such as from to part, macro micro, strategic tactical cultivate develop structured capability management.

Multi-resource arbitration is used where interchangeable resources need to be assigned among a number of client users. An efficient arbiter thus critical manage these shared resources. There has been only limited prior research on asynchronous solutions, which have mainly focused serial approaches. In this paper, new high-throughput multi-resource proposed, based fine-grain pipelining. The allocation resource divided into several steps, multiple successive client-resource pairs can selected...

The team-building of technology innovation talents is directly related to the survival and development enterprises.First this paper proposes an index system appraising enterprise talents'team, then use ANP(analytic network process) method analyze relation between indexes,and obtain weight each index. This reveals degree index's influencing talents' team, makes enterprises be clear about key factor, strives offer references for enterprises.