一种基于FCOS神经网络的小建筑物目标检测方法北大核心

提出了一种基于FCOS神经网络的小建筑物目标检测算法,针对FCOS模型在特征提取阶段提取到的小建筑物目标特征较少问题,引入多尺度检测和可变形卷积方式,加强网络对小建筑物目标的特征提取能力,并通过改进后的SGE注意力机制降低特征图中的干扰噪声权重。改进后的网络可以提取到更多的小建筑物目标特征,对环境干扰噪声的鲁棒性更强。在自己搭建的数据集上进行了实验测试,结果表明,在相同环境下网络改进后建筑物的整体检测准确率提升了1.7%,其中对小建筑物目标提升了3.6%,减少了小建筑物目标漏检、误检的问题。     

基于改进残差特征蒸馏的轻量级超分辨率网络

基于深度学习的图像超分辨率算法通常采用递归的方式或参数共享的策略来减少网络参数，这将增加网络的深度，使得运行网络花费大量的时间，从而很难将模型部署到现实生活中。为了解决上述问题，本文设计一种轻量级超分辨率网络，对中间特征的关联性及重要性进行学习，且在重建部分结合高分辨率图像的特征信息。首先，引入层间注意力模块，通过考虑层与层之间的相关性，自适应地分配重要层次特征的权重。其次，使用增强重建模块提取高分辨率图像中更精细的特征信息，以此得到更加清晰的重建图片。通过大量的对比实验表明，本文设计的网络与其他轻量级模型相比，有更小的网络参数量，并且在重建精度和视觉效果上都有一定的提升。     

Neural Emotion Detection via Personal Attributes

There has been a recent line of work to automatically detect the emotions of posts in social media. In literature, studies treat posts independently and detect their emotions separately. Different from previous studies, we explore the dependence among relevant posts via authors' backgrounds, since the authors with similar backgrounds, e.g., "gender", "location", tend to express similar emotions. However, personal attributes are not easy to obtain in most social media websites. Accordingly, we propose two approaches to determine personal attributes and capture personal attributes between different posts for emotion detection: the Joint Model with Personal Attention Mechanism (JPA) model is used to detect emotion and personal attributes jointly, and capture the attributes-aware words to connect similar people; the Neural Personal Discrimination (NPD) model is employed to determine the personal attributes from posts and connect the relevant posts with similar attributes for emotion detection. Experimental results show the usefulness of personal attributes in emotion detection, and the effectiveness of the proposed JPA and NPD approaches in capturing personal attributes over the state-of-the-art statistic and neural models.     

Korean automatic spacing using pretrained transformer encoder and analysis

Automatic spacing in Korean is used to correct spacing units in a given input sentence. The demand for automatic spacing has been increasing owing to frequent incorrect spacing in recent media, such as the Internet and mobile networks. Therefore, herein, we propose a transformer encoder that reads a sentence bidirectionally and can be pretrained using an out-of-task corpus. Notably, our model exhibited the highest character accuracy (98.42%) among the existing automatic spacing models for Korean. We experimentally validated the effectiveness of bidirectional encoding and pretraining for automatic spacing in Korean. Moreover, we conclude that pretraining is more important than fine-tuning and data size.     

Saliency4ASD: Challenge,dataset and tools for visual attention modeling for autism spectrum disorder

The recent studies showing that gaze features can be useful in the identification of Autism Spectrum Disorder (ASD), have opened a new domain where Visual Attention (VA) modeling could be of great help. In this sense, this paper presents a report of the Grand Challenge “Saliency4ASD: Visual attention modeling for Autism Spectrum Disorder”, organized at IEEE ICME’19, aiming at supporting the research on VA modeling towards this healthcare societal challenge. In particular, this paper describes the workflow, obtained results, and datasets and tools that were used within this activity, in order to help on the development and evaluation of two types of VA models: (1) to predict saliency maps that fit gaze behavior of people with ASD, and (2) to identify individuals with ASD from typical development.     

高地压严重突出区域下向穿层钻孔预抽消突技术

利用工作面高抽巷施工掩护顺槽掘进工作面下向穿层条带预抽钻孔,通过对钻场底板深、浅孔注浆封堵围岩裂隙解决了孔内返水不畅难题;采取了"两堵一注"封孔方式确保封孔严密不漏气;孔内全程下吹水管,实现了单孔、分组、定时吹水工作目标,提高了下向消突钻孔抽采效果,形成了一套高效、快速的下向穿层钻孔消突管理模式,实现了高地压严重突出区域煤层高效消突以及技术经济一体化目标,安全经济效益显著。     

结合注意力的3D卷积网络脑胶质瘤分割算法

为了提升脑胶质瘤分割精度，提出一种结合注意力机制的3D卷积神经网络算法。输入3个不同尺度的图像块，经过9个卷积层和1个分类层后得到3个不同的分类结果，将分类结果与注意力学习到的权重相乘并逐体素相加得到输出。此外该算法采用了一种混合Dice损失函数与Focal损失函数的超参数损失函数。实验表明，该算法的Dice系数在整体区域、核心区域以及增强区域分别达到了95.31%、80.12%、82.25%。与已有的一种脑胶质瘤分割算法deepmedic相比，整体区域、核心区域以及增强区域的Dice系数分别提升了3%、2%、6%。在脑胶质瘤分割方面，具有重要的临床意义。     

基于语言特征自动获取的反问句识别方法

反问句是以疑问的形式表达强烈情感的修辞方式,对其有效识别可为自然语言处理中的情感分析任务提供技术支持。该文提出了一种基于语言特征自动获取的反问句识别方法。首先,利用标签注意机制,建立了一个数据驱动的特征抽取模型,用于获取与任务相关的词汇、句法结构、符号标记和话题等语言特征。其次,利用Bi-LSTM模型分别对句子和语言特征进行表示,两者的交互注意被用于获取句子的各个词和符号的注意力权重向量。该权重向量作用于句子的表示,用于构建一个强化语言特征的反问句识别模型。在中文微博数据集上的实验结果表明,提出的方法与之前的工作相比,反问句识别性能有显著提升。     

A New Design Methodology for Time-Based Capacitance-to-Digital Converters (T-CDCs)

This paper introduces a 9-bit time-based capacitance-to-digital converter (T-CDC). This T-CDC adopts a new design methodology for parasitic cancellation with a simple calibration technique. In T-CDCs, the input sensor capacitance is first converted into a delay pulse using a capacitance-to-time converter (CTC) circuit; then this delay signal is converted into a digital code through a time-to-digital converter (TDC) circuit. A prototype of the proposed T-CDC is implemented in UMC 0.13 μm CMOS technology. This T-CDC consumes 8.42 μW and achieves a maximum SNR of 45.14 dB with a conversion time of 1 μs that corresponds to a figure of merit (FoM) of 16.4 fJ/Conv.     

结合句法依存信息的方面级情感分类

引入句法依存信息到原方面术语,提出一种新的方面术语表示方法,利用Glove词向量表示单词以及单词与单词之间的依存关系,构造出包含句法依存信息的依存关系邻接矩阵和依存关系表示矩阵,利用图卷积神经网络和多头注意力机制将句法依存信息融入到方面术语中,使得方面术语表达与上下文结构高度相关。将改进后的方面词术语表示替换到现有模型后,模型泛化能力得到有效提升。对比试验和分析结果表明:该方法具有有效性和泛化性。