Bayesian Neural Networks: An Introduction and Survey

Inferring the location of a user has been a valuable step for many applications that leverage social media, such as marketing, security monitoring and recommendation systems. Motivated by the recent success of Deep Learning techniques for many other tasks such as computer vision, speech recognition, and natural language processing, we study the application of neural networks to the problem of geolocation prediction and experiment with multiple techniques to improve neural networks for geolocation inference based solely on text. Experimental results on three Twitter datasets suggest that choosing appropriate network architecture, activation function, and performing Batch Normalization, can all increase performance on this task.

AraVec: A set of Arabic Word Embedding Models for use in Arabic NLP

For the past few years, gender classification has been an active area of study and researchers have been putting a lot of effort to contribute quality research in this area. There is a big potential field of study as it can be used in monitoring, surveillance and human-computer interaction. However, there is still a lack of the performance of existing methods on real live images. The rise of deep learning algorithm has been showing a spectacular increase in performance lately. Many difficult tasks involving computer vision, speech recognition, and natural language processing are easily solved with deep learning. Therefore, the approach to deep learning notably growing and this also happens to be on image classification. Gender classification is an important subject in the face recognition process. This paper shows the results of classifying gender using Convolutional Neural Network based Deep Learning architectures using Tensorflow’s Deep Learning framework. We have used models provided by Keras with weights pre-trained on ImageNet. We have made a comparison of the different type of models which includes VGG16, ResNet-50, and MobileNet. Our own database consists of Asian faces inclusive of Malaysians and some Caucasians. Our trained model on a database consisting of 1000 images shows that VGG-16 delivered the highest recognition accuracy.

Deep (machine) learning in recent years has significantly increased the predictive modeling strengths for applications in the areas of computer vision, speech recognition, and natural language processing. Being heavily data-driven, conventional general-purpose deep learning methods mainly rely on training data to learn task-specific data representations. In this thesis, we focus on leveraging novel learning algorithms in deep learning to tackle visual recognition challenges at the three key levels -- namely pixel, object, and image levels. Throughout the thesis, we demonstrate how each and every of our proposed methods can address a specific visual recognition problem. Firstly, we apply deep learning to improve the saliency detection performance of deep networks. A coarse saliency map of the input image's salient object is initially predicted by a convolutional-deconvolutional network (or equivalently a fully convolutional network). Following that, the coarse saliency map is sequentially refined by a separate network called Recurrent Attentional Convolutional-Deconvolutional Network (RACDNN) that shares its network weights across its sequential steps. At every sequential (temporal) step, the same RACDNN is applied to a newly selected/attended image sub-region and it performs saliency map refinement for that particular sub-region. Despite having many learnable network parameters and no access to additional training data, RADCNN can be trained well and it outperforms the "single-step" (coarse saliency detection) baseline and state-of-the-art saliency detection methods by large margins, on several saliency detection benchmark datasets. Secondly, we introduce auto-encoder weight transfer networks (AE-WTN+), a deep model for scaling up object detection. Conventional WTN transfers object class knowledge from the classification weights of a pretrained large-scale image classification network, to the classification weights of an object detection network. However, they suffer from under-fitting and are not able to preserve the rich object class information of novel classes. In this work, we propose a new WTN method -- AE-WTN+ which makes use of normalization techniques to ease network optimization. Inputs and intermediate features are normalized to improve trainability for reduced under-fitting. More importantly, a novel auto-encoding (reconstruction) loss is proposed to encourage information preservation of all classes in the weight predictions of AE-WTN+. This enables the predicted weights of AE-WTN+ to maintain the class discriminability and neighbourhood relationships contained in the pre-trained large-scale classification weights. Experiments on large-scale detection datasets validate the effectiveness of AE-WTN+ on both seen and novel classes. Thirdly, we improve the efficiency of cross-layer network connections for very deep convolutional network in the context of image classification. This gives rise to a new convolutional network architecture called Deluge Networks (DelugeNets) which pass the output (feature activations) of any preceding composite network layer as direct inputs to all succeeding composite layers. For the incoming feature activations of preceding composite layers, the succeeding composite layers first apply rather lightweight cross-layer depthwise convolutional weights on them for feature aggregation. The heavy lifting is carried out by the last convolutional weight layer in each preceding composite layer that performs cross-channel convolutions. The proposed technique enables unimpeded information flow among many layers in very deep DelugeNets with great efficiency. Experiments on small-scale and large-scale datasets show that DelugeNets perform competitively to state-of-the-art deep convolutional networks at lower computational and parameter costs. Fourthly, we propose Stochastic Downsampling (SDPoint), a method to overcome a major limitation suffered by conventional deep convolutional networks -- their computational costs during inference are fixed and cannot be changed. SDPoint downsamples the feature maps at a randomly selected point (layer index) in the network hierarchy during training. The many downsampling configurations known as SDPoint instances each entail a unique downsampling point and downsampling ratio. The same convolutional network weights are shared across the different SDPoint instances that downsample feature maps differently. During inference, the computational cost of a network trained with SDPoint can be instantaneously adjusted to fit a given computational budget, by manually selecting an appropriate SDPoint instance. Additionally, sharing weights between different feature map scales provides significant regularization benefits, making the networks trained with SDPoint less sensitive to scales and more competent. In image classification experiments, SDPoint demonstrates significant cost-accuracy and computational advantages, over independently-trained (non-weight-shared) baselines with various inference costs.

Recently, significant improvements have been made in artificial intelligence. The artificial neural network was introduced in the 1950s. However, because of the low computing power and insufficient datasets available at that time, artificial neural networks suffered from overfitting and vanishing gradient problems for training deep networks. This concept has become more promising owing to the enhanced big data processing capability, improvement in computing power with parallel processing units, and new algorithms for deep neural networks, which are becoming increasingly successful and attracting interest in many domains, including computer vision, speech recognition, and natural language processing. Recent studies in this technology augur well for medical and healthcare applications, especially in endoscopic imaging. This paper provides perspectives on the history, development, applications, and challenges of deep-learning technology.

Nowadays, deep learning has been widely used in various fields, and great breakthroughs have been made in computer vision, speech recognition and natural language processing. Based on the deep learning algorithm, this paper studies the general optimization technology of OpenCL based on FPGA in detail, and puts forward a targeted hardware design scheme, which has the characteristics of high performance, low power consumption and programmability. By analyzing the algorithm of forward calculation and reverse calculation in convolutional neural network, the hardware acceleration module is designed. Compared with CPU, GPU and other platforms, FPGA has obvious advantages in power consumption, which proves that the scheme proposed in this paper has good feasibility and accuracy.

Learning data representations directly from the data itself is an approach that has shown great success in different pattern recognition problems, outperforming state-of-the-art feature extraction schemes for different tasks in computer vision, speech recognition and natural language processing. Representation learning applies unsupervised and supervised machine learning methods to large amounts of data to find building-blocks that better represent the information in it. Digitized histopathology images represents a very good testbed for representation learning since it involves large amounts of high complex, visual data. This paper presents a comparative evaluation of different supervised and unsupervised representation learning architectures to specifically address open questions on what type of learning architectures (deep or shallow), type of learning (unsupervised or supervised) is optimal. In this paper we limit ourselves to addressing these questions in the context of distinguishing between anaplastic and non-anaplastic medulloblastomas from routine haematoxylin and eosin stained images. The unsupervised approaches evaluated were sparse autoencoders and topographic reconstruct independent component analysis, and the supervised approach was convolutional neural networks. Experimental results show that shallow architectures with more neurons are better than deeper architectures without taking into account local space invariances and that topographic constraints provide useful invariant features in scale and rotations for efficient tumor differentiation.

Artificial Intelligence (AI) has become one of the most transformative foundational technologies of our time cutting across industry and function, while machine learning has become its engine of maturation. The rapid advancement of AI applications in different domains including computer vision, speech recognition, and natural language processing (NLP) has led to a need for highly accurate, efficient, scalable, and deployable systems in the real world. The paper discusses advances needed to advance AI applicability in these three major areas, especially using advanced machine-learning technologies to improve AI efficiency in terms of the performance, energy requirements, and computational scalability of AI systems In order to seek out the accuracy and speed of the computer vision systems, the study’s investigation studies the works related to the modernization of machine learning techniques namely deep learning, reinforcement learning, and transfer learning. Such advances are critical for real-time applications including ampere nous driving, facial recognizance, and medical image diagnosis. In parallel, studies on the architecture optimization for speech recognition models are being conducted to enhance feature extraction, reduce computation and latency while maintaining performance. Such improvements are crucial for voice assistants, transcription services and multilingual communication tools. For NLP, these translations shed light on the latest in transformer-based architectures, better pre-trained language models, and methods for reducing the size and inference time of models, all so that these AI services can be reasonable in chatbots, content summarization, and sentiment analysis, etc. This work presents a detailed insight into how machine learning techniques can offer massive energy savings that can be derived from the analysis of state-of-the-art architectures, finding bottlenecks, and comparing performance. In addition, it underscores recent advances in lightweight models, distributed training, model compression and hardware-aware optimization engage computational burden while preserving or improving predictive performance. This brings these approaches together to maintain that AI technologies can be consolidated more sustainably across edge devices, cloud platforms, and embedded systems.

Great progress has been made in deep learning over the past few years, which drives the deployment of deep learning–based applications into cyber-physical systems. But the lack of interpretability for deep learning models has led to potential security holes. Recent research has found that deep neural networks are vulnerable to well-designed input examples, called adversarial examples . Such examples are often too small to detect, but they completely fool deep learning models. In practice, adversarial attacks pose a serious threat to the success of deep learning. With the continuous development of deep learning applications, adversarial examples for different fields have also received attention. In this article, we summarize the methods of generating adversarial examples in computer vision, speech recognition, and natural language processing and study the applications of adversarial examples. We also explore emerging research and open problems.

Deep neural networks (DNNs) have transformed several artificial intelligence research areas including computer vision, speech recognition, and natural language processing. However, recent studies demonstrated that DNNs are vulnerable to adversarial manipulations at testing time. Specifically, suppose we have a testing example, whose label can be correctly predicted by a DNN classifier. An attacker can add a small carefully crafted noise to the testing example such that the DNN classifier predicts an incorrect label, where the crafted testing example is called adversarial example. Such attacks are called evasion attacks. Evasion attacks are one of the biggest challenges for deploying DNNs in safety and security critical applications such as self-driving cars.

Various machine learning architectures including neural networks have been designed, developed and used to classify data. These networks have been used for Computer Vision, Speech Recognition and Natural Language Processing, to mention but a few and provide near accurate results. One of the major challenges faced in the area of mathematical equational recognition has been background information and noise. This paper presents a system that makes use of image processing and an artificial neural network to recognize, contextualize and compute mathematical equations from noisy images. The system attempts to overcome the challenges faced at segmentation and recognition stages.

Microfluidic-based organs-on-chips (OoCs) are a rapidly developing technology in biomedical and chemical research and have emerged as one of the most advanced and promising in vitro models. The miniaturization, stimulated tissue mechanical forces, and microenvironment of OoCs offer unique properties for biomedical applications. However, the large amount of data generated by the high parallelization of OoC systems has grown far beyond the scope of manual analysis by researchers with biomedical backgrounds. Deep learning, an emerging area of research in the field of machine learning, can automatically mine the inherent characteristics and laws of “big data” and has achieved remarkable applications in computer vision, speech recognition, and natural language processing. The integration of deep learning in OoCs is an emerging field that holds enormous potential for drug development, disease modeling, and personalized medicine. This review briefly describes the basic concepts and mechanisms of microfluidics and deep learning and summarizes their successful integration. We then analyze the combination of OoCs and deep learning for image digitization, data analysis, and automation. Finally, the problems faced in current applications are discussed, and future perspectives and suggestions are provided to further strengthen this integration.

Learning a similarity function between pairs of objects is at the core of learning to rank approaches. In information retrieval tasks we typically deal with query-document pairs, in question answering -- question-answer pairs. However, before learning can take place, such pairs needs to be mapped from the original space of symbolic words into some feature space encoding various aspects of their relatedness, e.g. lexical, syntactic and semantic. Feature engineering is often a laborious task and may require external knowledge sources that are not always available or difficult to obtain. Recently, deep learning approaches have gained a lot of attention from the research community and industry for their ability to automatically learn optimal feature representation for a given task, while claiming state-of-the-art performance in many tasks in computer vision, speech recognition and natural language processing. In this paper, we present a convolutional neural network architecture for reranking pairs of short texts, where we learn the optimal representation of text pairs and a similarity function to relate them in a supervised way from the available training data. Our network takes only words in the input, thus requiring minimal preprocessing. In particular, we consider the task of reranking short text pairs where elements of the pair are sentences. We test our deep learning system on two popular retrieval tasks from TREC: Question Answering and Microblog Retrieval. Our model demonstrates strong performance on the first task beating previous state-of-the-art systems by about 3\% absolute points in both MAP and MRR and shows comparable results on tweet reranking, while enjoying the benefits of no manual feature engineering and no additional syntactic parsers.

As an important technology and research direction to achieve AI, deep learning has been widely applied in the fields of computer vision, speech recognition, and natural language processing. How to effectively accelerate the computing power of deep learning has always been the focus of scientific research. Among various acceleration technologies, FPGA has the advantages of reconfiguration, high performance, small size, and low latency. As more and more FPGA-based neural network accelerators are developed, we notice there is a lack of a complete and detailed overview. In this paper, we give a comparative study of DNN accelerators on FPGA from the aspects of hardware structures, design ideas and optimization strategies. We further compare the performance of different acceleration technologies in different models and present the prospects of the FPGA accelerators for deep learning.

Convolutional neural networks (CNNs) have demonstrated promising results in various applications such as computer vision, speech recognition, and natural language processing. One of the key computations in many CNN applications is matrix multiplication, which accounts for a significant portion of computation. Therefore, hardware accelerators to effectively speed up the computation of matrix multiplication have been proposed, and several studies have attempted to design hardware accelerators to perform better matrix multiplications in terms of both speed and power consumption. Typically, accelerators with either a two-dimensional (2D) systolic array structure or a single instruction multiple data (SIMD) architecture are effective only when the input matrix has shapes that are close to or similar to a square. However, several CNN applications require multiplications of non-squared matrices with various shapes and dimensions, and such irregular shapes lead to poor utilization efficiency of the processing elements (PEs). This study proposes a configurable engine for neural network acceleration, called CONNA, whose computation engine can conduct matrix multiplications with highly utilized computing units, regardless of the access patterns, shapes, and dimensions of the input matrices by changing the shape of matrix multiplication conducted in the physical array. To verify the functionality of the CONNA accelerator, we implemented CONNA as an SoC platform that integrates a RISC-V MCU with CONNA on Xilinx VC707 FPGA. SqueezeNet on CONNA achieved an inference performance of 100 frames per second (FPS) with 2.36 mm2 and 83.55 mW in a 65 nm process, improving efficiency by up to 34.1 times better than existing accelerators in terms of FPS, silicon area, and power consumption.