FPGA-based implementation of classification techniques: A survey

Recently, a number of classification techniques have been introduced. However, processing large dataset in a reasonable time has become a major challenge. This made classification task more complex and expensive in calculation. Thus, the need for solutions to overcome these constraints such as field programmable gate arrays (FPGAs). In this paper, we give an overview of the various classification techniques. Then, we present the existing FPGA based implementation of these classification methods. After that, we investigate the confronted challenges and the optimizations strategies. Finally, we highlight the hardware accelerator architectures and tools for hardware design suggested to improve the FPGA implementation of classification methods.     

Residual attention-based tracking-by-detection network with attention-driven data augmentation

Tracking-by-detection (TBD) is a significant framework for visual object tracking. However, current trackers are usually updated online based on random sampling with a probability distribution. The performance of the learning-based TBD trackers is limited by the lack of discriminative features, especially when the background is full of semantic distractors. We propose an attention-driven data augmentation method, in which a residual attention mechanism is integrated into the TBD tracking network as supplementary references to identify discriminative image features. A mask generating network is used to simulate changes in target appearances to obtain positive samples, where attention information and image features are combined to identify discriminative features. In addition, we propose a method for mining hard negative samples, which searches for semantic distractors with the response of the attention module. The experiments on the OTB2015, UAV123, and LaSOT benchmarks show that this method achieves competitive performance in terms of accuracy and robustness.     

Structural Aspects and Prediction of Calmodulin-Binding Proteins

Calmodulin (CaM) is an important intracellular protein that binds Ca²⁺ and functions as a critical second messenger involved in numerous biological activities through extensive interactions with proteins and peptides. CaM’s ability to adapt to binding targets with different structures is related to the flexible central helix separating the N- and C-terminal lobes, which allows for conformational changes between extended and collapsed forms of the protein. CaM-binding targets are most often identified using prediction algorithms that utilize sequence and structural data to predict regions of peptides and proteins that can interact with CaM. In this review, we provide an overview of different CaM-binding proteins, the motifs through which they interact with CaM, and shared properties that make them good binding partners for CaM. Additionally, we discuss the historical and current methods for predicting CaM binding, and the similarities and differences between these methods and their relative success at prediction. As new CaM-binding proteins are identified and classified, we will gain a broader understanding of the biological processes regulated through changes in Ca²⁺ concentration through interactions with CaM.     

Vulnerable objects detection for autonomous driving: A review

Object detection performed by Autonomous Vehicles (AV)s is a crucial operation that comes ahead of various autonomous driving tasks, such as object tracking, trajectories estimation, and collision avoidance. Dynamic road elements (pedestrians, cyclists, vehicles) impose a greater challenge due to their continuously changing location and behaviour. This paper presents a comprehensive review of the state-of-the-art object detection technologies focusing on both the sensory systems and algorithms used. It begins with a brief introduction on the autonomous driving operations and challenges. Then, different sensory systems employed on existing AVs are elaborated while illustrating their advantages, limitations and applications. Also, sensory systems employed by different research are reviewed. Moreover, due to the significant role Deep Neural Networks (DNN)s are playing in object detection tasks, different DNN-based networks are also highlighted. Afterwards, previous research on dynamic objects detection performed by AVs are reviewed in tabular forms. Finally, a conclusion summarizes the outcomes of the review and suggests future work towards the development of vehicles with higher automation levels.     

Technology enhanced learning of quantitative critical thinking

A Quantitative Critical Thinking (QCT) software tool was developed in this study to facilitate students’ learning of quantitative critical thinking via repeated practice by chemical engineering students reading a core module called fluid-solid systems. The software tool generated detailed calculation steps to typical engineering design problems encountered in this module that contained weaknesses, flaws or even errors. Students utilized the software tool to practice identifying these weaknesses, flaws or errors in the design solutions and then present a better or correct design by applying the concepts and knowledge acquired in the module. Since the QCT software tool was built upon an existing design software tool that was able to generate the correct, detailed design calculation steps to design problems, students were able to check their own design calculations against those presented by the software tool during this second learning step, thereby engaging in and learning quantitative critical thinking via a repeated practice approach. The software tool was successful in enhancing the performance of second-year undergraduate students in solving a question that required quantitative critical thinking in the final examination of the module. The average percentage scores achieved by students for the question who reported higher frequencies of usage of the software were generally higher than those who reported lower frequencies of usage or did not utilize the software tool throughout the semester.     

工艺参数对动态流量控制法挤压弯管应力的影响

采用有限元模拟和实验研究了挤压钛合金弯曲管件。通过实验验证了工件的形状和尺寸精度,并通过有限元模拟分析了工艺参数对挤出过程中变形体的平均压应力分布情况和挤出弯管件的曲率半径的影响规律。结果表明:有限元模拟中,弯管件的曲率半径误差为6.03%,弯管直径误差为3.82%;在靠近定径带处,平均压应力呈非均匀分布;在焊合腔内,靠近细分流孔区域的平均压应力小于靠近粗分流孔区域的平均压应力,平均压应力的大小顺序在通过粗、细分流孔前后相反;在模具结构固定不变时,弯管件的曲率半径随挤压速度的减小而增大,不随挤压温度的变化而变化。     

梯度铝蜂窝夹芯板的力学行为

梯度分层铝合金蜂窝板是一种有效的吸能结构,本工作在梯度铝蜂窝结构的基础上根据梯度率的概念,通过改变蜂窝芯层的胞壁长度,设计了4种质量相同、梯度率不同的铝蜂窝夹芯结构。通过准静态压缩实验,并结合非线性有限元模拟准静态及冲击态下梯度铝蜂窝夹芯结构的变形情况及其力学性能,分析对比了相同质量下梯度铝蜂窝夹芯结构在准静态下的变形模式以及冲击载荷下分层均质蜂窝结构和不同梯度率的分层梯度蜂窝结构的动态响应和能量吸收特性。结果表明:在准静态压缩过程中,铝蜂窝梯度夹芯板的变形具有明显的局部化特征,蜂窝芯的变形为低密度优先变形直至密实,层级之间的密实化应变差随芯层密度的增大而逐渐减小;在高速冲击下,梯度蜂窝板并非严格按照准静态过程中逐级变形直至密实,而是在锤头冲击惯性及芯层密度的相互作用下整体发生的线弹性变形、弹性屈曲、塑性坍塌及密实化;另外,在本工作所设计的梯度率中,当梯度率为γ₁=0.0276时,梯度蜂窝夹芯板的吸能性达到最好,相较于同等质量下的均质蜂窝夹芯板,能量吸收提高了10.63%。     

Past is important: Improved image captioning by looking back in time

A major development in the area of image captioning consists of trying to incorporate visual attention in the design of language generative model. However, most previous studies only emphasize its role in enhancing visual composition at the current moment, while neglect its role in global sequence reasoning. This problem appears not only in captioning model, but also in reinforcement learning structure. To tackle this issue, we first propose a Visual Reserved model that enables previous visual context to be considered for the current sequence reasoning. Next, a Attentional-Fluctuation Supervised model is also proposed in reinforcement learning structure. Compared against the traditional strategies that only take non-differentiable Natural Language Processing (NLP) metrics as the incentive standard, the proposed model regards the fluctuation of previous attention matrix as an important indicator to judge the convergence of the captioning model. The proposed methods have been tested on MS-COCO captioning dataset and achieve competitive results evaluated by the evaluation server of MS COCO captioning challenge.     

Molecular fingerprint and machine learning to accelerate design of high-performance homochiral metal–organic frameworks

Computational screening was employed to calculate the enantioseparation capabilities of 45 functionalized homochiral metal–organic frameworks (FHMOFs), and machine learning (ML) and molecular fingerprint (MF) techniques were used to find new FHMOFs with high performance. With increasing temperature, the enantioselectivities for (R,S)-1,3-dimethyl-1,2-propadiene are improved. The “glove effect” in the chiral pockets was proposed to explain the correlations between the steric effect of functional groups and performance of FHMOFs. Moreover, the neighborhood component analysis and RDKit/MACCS MFs show the highest predictive effect on enantioselectivities among the four ML classification algorithms with nine MFs that were tested. Based on the importance of MF, 85 new FHMOFs were designed, and a newly designed FHMOF, NO₂-NHOH-FHMOF, with high similarity to the optimal MFs achieved improved chiral separation performance, with enantioselectivities of 85%. The design principles and new chiral pockets obtained by ML and MFs could facilitate the development of new materials for chiral separation.