A hybrid neural network hidden Markov model approach for automatic story segmentation

This paper provides an extended comparison of two temporal models for gesture recognition, namely Hybrid Neural Network-Hidden Markov Models (NN-HMM) and Recurrent Neural Networks (RNN) which have lately claimed the state-the-art performances. Experiments were conducted on both models in the same body of work, with similar representation learning capacity and comparable computational costs. For both solutions, we have integrated recent contributions to the model architectures and training techniques. We show that, for this task, Hybrid NN-HMM models remain competitive with Recurrent Neural Networks in a standard setting. For both models, we analyze the influence of the training objective function on the final evaluation metric. We further tested the influence of temporal convolution to improve context modeling, a technique which was recently reported to improve the accuracy of gesture recognition.

This paper proposes to select frame-sized speech segments for waveform concatenation speech synthesis using neural network based acoustic models. First, a deep neural network (DNN) based frame selection method is presented. In this method, three DNNs are adopted to calculate target costs and concatenation costs respectively for selecting candidate frames of 5ms length. One DNN is built in the same way as the DNN-based statistical parametric speech synthesis, which predicts target acoustic features given linguistic context inputs. The distance between the acoustic features of a candidate unit and the predicted ones for a target unit is calculated as the target cost. The other two DNNs are constructed to predict the acoustic features at current frame using its context features and the acoustic features of preceding frames. At synthesis time, these two DNNs are employed to calculate the concatenation cost for each candidate frame given its preceding ones. Furthermore, recurrent neural networks (RNNs) with long short-term memory (LSTM) cells are adopted to replace DNNs for acoustic modeling in order to make better use of the sequential information. A strategy of using multi-frame instead of single frame as the basic unit for selection is also presented to reduce the concatenation points within synthetic speech. Experimental results show that our proposed method can achieve better naturalness than the hidden Markov model (HMM)-based frame selection method and the HMM-based parametric speech synthesis method.

Besides the feedforward neural networks, there are the recurrent networks, where the impulses can be transmitted in both directions due to some reaction connections in these networks. Recurrent Neural Networks (RNNs) are linear or nonlinear dynamic systems. The dynamic behavior presented by the recurrent neural networks can be described both in continuous time, by differential equations and at discrete times by the recurrence relations (difference equations). The distinction between recurrent (or dynamic) neural networks and static neural networks is due to recurrent connections both between the layers of neurons of these networks and within the same layer, too. The aim of this chapter is to describe a Recurrent Fuzzy Neural Network (RFNN) model, whose learning algorithm is based on the Improved Particle Swarm Optimization (IPSO) method.

<p>Deep neural networks have achieved tremendous success in many domains (e.g., computer vision~\cite{Alexnet12,vggnet15,fastrcnn15}, speech recognition~\cite{hinton2012deep,dahl2012context}, natural language processing~\cite{dahl2012context,collobert2011natural}, games~\cite{silver2017mastering,silver2016mastering}), however, there are still many challenges in deep learning comunity such as how to speed up training large deep neural networks, how to compress large nerual networks for mobile/embed device without performance loss, how to automatically design the optimal network structures for a certain task, and how to further design the optimal networks with improved performance and certain model size with reduced computation cost.</p> <p>To speed up training large neural networks, we propose to use multinomial sampling for dropout, i.e., sampling features or neurons according to a multinomial distribution with different probabilities for different features/neurons. To exhibit the optimal dropout probabilities, we analyze the shallow learning with multinomial dropout and establish the risk bound for stochastic optimization. By minimizing a sampling dependent factor in the risk bound, we obtain a distribution-dependent dropout with sampling probabilities dependent on the second order statistics of the data distribution. To tackle the issue of evolving distribution of neurons in deep learning, we propose an efficient adaptive dropout (named evolutional dropout) that computes the sampling probabilities on-the-fly from a mini-batch of examples.</p> <p>To compress large neural network structures, we propose a simple yet powerful method for compressing the size of deep Convolutional Neural Networks (CNNs) based on parameter binarization. The striking difference from most previous work on parameter binarization/quantization lies at different treatments of $1\times 1$ convolutions and $k\times k$ convolutions ($k>1$), where we only binarize $k\times k$ convolutions into binary patterns. By doing this, we show that previous deep CNNs such as GoogLeNet and Inception-type Nets can be compressed dramatically with marginal drop in performance. Second, in light of the different functionalities of $1\times 1$ (data projection/transformation) and $k\times k$ convolutions (pattern extraction), we propose a new block structure codenamed the pattern residual block that adds transformed feature maps generated by $1\times 1$ convolutions to the pattern feature maps generated by $k\times k$ convolutions, based on which we design a small network with $\sim 1$ million parameters. Combining with our parameter binarization, we achieve better performance on ImageNet than using similar sized networks including recently released Google MobileNets.</p> <p>To automatically design neural networks, we study how to design a genetic programming approach for optimizing the structure of a CNN for a given task under limited computational resources yet without imposing strong restrictions on the search space. To reduce the computational costs, we propose two general strategies that are observed to be helpful: (i) aggressively selecting strongest individuals for survival and reproduction, and killing weaker individuals at a very early age; (ii) increasing mutation frequency to encourage diversity and faster evolution. The combined strategy with additional optimization techniques allows us to explore a large search space but with affordable computational costs.</p> <p>To further design the optimal networks with improved performance and certain model size under reduced computation cost, we propose an ecologically inspired genetic approach for neural network structure search , that includes two types of succession: primary and secondary succession as well as accelerated extinction. Specifically, we first use primary succession to rapidly evolve a community of poor initialized neural network structures into a more diverse community, followed by a secondary succession stage for fine-grained searching based on the networks from the primary succession. Accelerated extinction is applied in both stages to reduce computational cost. In addition, we also introduce the gene duplication to further utilize the novel block of layers that appeared in the discovered network structure.</p>

Deep Neural Networks (DNNs) have been recently in the focus of interest owing to their high-quality results in various application domains of big data analytics. The lack of end-to-end toolchains that can automatically optimize and translate a DNN algorithm, usually implemented in high-level languages, on FPGA-based platform hinders software developers to exploit such platforms. A few available toolchains, namely Xilinx DPU, have been optimized for computational expensive applications, e.g., real-time image and video processing, and therefore they introduce large performance overhead on lightweight applications, e.g., sensor-driven decision analytics. In this paper, we introduce the FLight framework to fully automatize the acceleration of lightweight DNN algorithms on FPGA-based embedded platforms. The framework takes a trained model in TensorFlow/Keras as input and generates an optimized synthesizable C++ version and maps it on the target platform using HLS tools. FLight is an easy-to-use framework and does not require deep knowledge of embedded system design since it automatically performs all the required mapping steps. FLight supports the acceleration of various deep learning algorithms, e.g., Feed-Forward Neural networks (FFNNs), Convolutional Neural Networks (CNNs), and sequential models like Recurrent Neural Networks (RNNs). We evaluated the applicability of FLight with various DNN models on both academic and industrial datasets. As a case study, we exploited the FLight framework for wind turbine condition monitoring using an industrial dataset provided by Weidmüller Monitoring Systems GmbH. The experimental results revealed the 40 times speedup compared to the Xilinx DPU framework.

The Sixth International Symposium on Neural Networks (ISNN 2009)

Traffic prediction plays an important role in the intelligent transportation system (ITS), because it can increase people’s travel convenience. Despite the deep neural network has been widely used in the field of traffic prediction, literature surveys of such methods and data categories are rare. In this paper, we have a summary of traffic forecasting from data, methods and challenges. Firstly, we are according to the difference of in spatio-temporal dimensions, divide the data into three types, including the spatio-temporal static data, spatial static time dynamic data, and spatio-temporal dynamic data. Secondly, we explore three significant neural networks of deep learning in traffic prediction, including the convolutional neural network (CNN), the recurrent neural network (RNN), and the hybrid neural networks models. These methods are used in many aspects of traffic prediction, including road traffic accidents forecast, road traffic flow prediction, road traffic speed forecast, and road traffic congestion forecast introduced. Finally, we provide a discussion of some current challenges and development prospects. KeywordsDeep neural networkTraffic forecastingSpatio-temporal data

Determining the structure of a protein given its sequence is a challenging problem. Deep learning is a rapidly evolving field which excels at problems where there are complex relationships between input features and desired outputs. Deep Neural Networks have become popular for solving problems in protein science. Various deep neural network architectures have been proposed including deep feed-forward neural networks, recurrent neural networks and more recently neural Turing machines and memory networks. This article provides a short review of deep learning applied to protein prediction problems.

The adoption of wind energy has grown significantly in recent years. New, cost-effective technologies have been developed, led by customer awareness of green technologies and a legal framework proposed at the European Union level. The stochastic nature of wind speed is transferred to wind turbine output, making wind energy difficult to predict. The main scope of predicting wind energy production is to be proactive in balancing and reserving energy to meet demand. When the prediction identifies a potential gap between supply and demand, additional energy from other sources must be generated and supplied. Creating a synergy of physical devices through advanced sensing capabilities, software, storage and analytics capabilities, the Industrial Internet of Things is enabling the effective transition to wind energy through automation by removing many of the disadvantages in a way that has recently become accessible. This research focuses on the data analytics, proposing a fast univariate network-based approach for wind energy prediction, using Feed Forward Neural Networks, Recurrent Neural Networks, Long-Short Term Memory, Gated Recurrent Unit, and Convolutional Neural Networks. Moreover, by introducing the theoretical fundamentals, the implementation method and the hyperparameters of the final models, this article becomes unique in the context of wind energy. At the time of this study, no prior research studies have presented a direct comparison between feedforward, recurrent, and convolutional neural networks ? these being the most important in the field of supervised learning.

Deep neural networks (DNNs) are increasingly expanding their real-world applications across domains, e.g., image processing, speech recognition and natural language processing. However, there is still limited tool support for DNN testing in terms of test data quality and model robustness. In this paper, we introduce a mutation testing-based tool for DNNs, DeepMutation++, which facilitates the DNN quality evaluation, supporting both feed-forward neural networks (FNNs) and stateful recurrent neural networks (RNNs). It not only enables static analysis of the robustness of a DNN model against the input as a whole, but also allows the identification of the vulnerable segments of a sequential input (e.g. audio input) by runtime analysis. It is worth noting that DeepMutation++ specially features the support of RNNs mutation testing. The tool demo video can be found on the project website https://sites.google.com/view/deepmutationpp.

Stacked autoencoder (SAE) is hard to achieve satisfactory performance, when input data are complex and non-stationary. Besides, the identification performance of recurrent neural network (RNN) may decrease rapidly under noisy environment. In order to deal with these problems, a novel hybrid deep neural network (DNN) based on stacked denoising autoencoder (SDAE) and gated recurrent unit neural network (GRUNN) is presented. First, the structure of the presented hybrid DNN is given. The hybrid DNN contains a SDAE, a GRUNN, and a softmax classifier. Then, the training algorithm based on action discovery (AD) is proposed to train the presented hybrid DNN. The experimental studies indicate the presented hybrid DNN processes strong anti-noise ability and adaptability to time-varying signals.

We propose an end-to-end approach for modeling polyphonic music with a novel graphical representation, based on music theory, in a deep neural network. Despite the success of deep learning in various applications, it remains a challenge to incorporate existing domain knowledge in a network without affecting its training routines. In this paper we present a novel approach for predictive music modeling and music generation that incorporates domain knowledge in its representation. In this work, music is transformed into a 2D representation, inspired by tonnetz from music theory, which graphically encodes musical relationships between pitches. This representation is incorporated in a deep network structure consisting of multilayered convolutional neural networks (CNN, for learning an efficient abstract encoding of the representation) and recurrent neural networks with long short-term memory cells (LSTM, for capturing temporal dependencies in music sequences). We empirically evaluate the nature and the effectiveness of the network by using a dataset of classical music from various composers. We investigate the effect of parameters including the number of convolution feature maps, pooling strategies, and three configurations of the network: LSTM without CNN, LSTM with CNN (pre-trained vs. not pre-trained). Visualizations of the feature maps and filters in the CNN are explored, and a comparison is made between the proposed tonnetz-inspired representation and pianoroll, a commonly used representation of music in computational systems. Experimental results show that the tonnetz representation produces musical sequences that are more tonally stable and contain more repeated patterns than sequences generated by pianoroll-based models, a finding that is directly useful for tackling current challenges in music and AI such as smart music generation.

The article deals with the problem of path planning in a two-dimensional environment based on deep learning neural networks. Deep neural networks require large amounts of data and place high computational requirements on computing tools. The lack of sufficient data leads to a decrease in the accuracy of the neural network, and high computational requirements at the learning stage limit the use of this technology in engineering practice. In this paper, the forms of representation of the environment for the input of a neural network are studied. Vector form allows to reduce the amount of information supplied to the input of a neural network, but it leads to the need to use more complex neural networks. In this article, a combined form of representation is proposed, including a vector global and local map layout. The vector part of the map includes the position of the robot, the position of the target point and a description of obstacles. The local raster map describes the area closest to the robot. Using numerical research, the effectiveness of this form of data representation for a precise neural network is shown, compared with the raster representation of the map. In this article, two structures of neural networks are studied, one of which uses 8 possible directions of movement, and the other uses 3 possible directions of movement. It is shown that when using 3 possible directions, the cycling of trajectories planned by the neural network is eliminated, which leads to an increase in accuracy.

Feature data processing: Making medical data fit deep neural networks

Deep neural networks (DNNs) have shown a great achievement in acoustic modeling for speech recognition task. Of these networks, convolutional neural network (CNN) is an effective network for representing the local properties of the speech formants. However, CNN is not suitable for modeling the long-term context dependencies between speech signal frames. Recently, the recurrent neural networks (RNNs) have shown great abilities for modeling long-term context dependencies. However, the performance of RNNs is not good for low-resource speech recognition tasks, and is even worse than the conventional feed-forward neural networks. Moreover, these networks often overfit severely on the training corpus in the low-resource speech recognition tasks. This paper presents the results of our contributions to combine CNN and conventional RNN with gate, highway, and residual networks to reduce the above problems. The optimal neural network structures and training strategies for the proposed neural network models are explored. Experiments were conducted on the Amharic and Chaha datasets, as well as on the limited language packages (10-h) of the benchmark datasets released under the Intelligence Advanced Research Projects Activity (IARPA) Babel Program. The proposed neural network models achieve 0.1–42.79% relative performance improvements over their corresponding feed-forward DNN, CNN, bidirectional RNN (BRNN), or bidirectional gated recurrent unit (BGRU) baselines across six language collections. These approaches are promising candidates for developing better performance acoustic models for low-resource speech recognition tasks.