https://bura.brunel.ac.uk/bitstream/2438/14221/1/FullText.pdf

A survey of deep neural network architectures and their applications

An intelligent automated assistant system engages with the user in an integrated, conversational manner using natural language dialog, and invokes external services when appropriate to obtain information or perform various actions. The system can be implemented using any of a number of different platforms, such as the web, email, smartphone, and the like, or any combination thereof. In one embodiment, the system is based on sets of interrelated domains and tasks, and employs additional functionally powered by external services with which the system can interact.

Intelligent Automated Assistant

Also referred to as learning by imitation, tutelage, or apprenticeship learning, Programming by Demonstration (PbD) develops methods by which new skills can be transmitted to a robot. This book examines methods by which robots learn new skills through human guidance. Taking a practical perspective, it covers a broad range of applications, including service robots. The text addresses the challenges involved in investigating methods by which PbD is used to provide robots with a generic and adaptive model of control. Drawing on findings from robot control, human-robot interaction, applied machine learning, artificial intelligence, and developmental and cognitive psychology, the book contains a large set of didactic and illustrative examples. Practical and comprehensive machine learning source codes are available on the books companion website: http://www.programming-by-demonstration.org

Robot Programming by Demonstration

Deep learning (DL) is playing an increasingly important role in our lives. It has already made a huge impact in areas, such as cancer diagnosis, precision medicine, self-driving cars, predictive forecasting, and speech recognition. The painstakingly handcrafted feature extractors used in traditional learning, classification, and pattern recognition systems are not scalable for large-sized data sets. In many cases, depending on the problem complexity, DL can also overcome the limitations of earlier shallow networks that prevented efficient training and abstractions of hierarchical representations of multi-dimensional training data. Deep neural network (DNN) uses multiple (deep) layers of units with highly optimized algorithms and architectures. This paper reviews several optimization methods to improve the accuracy of the training and to reduce training time. We delve into the math behind training algorithms used in recent deep networks. We describe current shortcomings, enhancements, and implementations. The review also covers different types of deep architectures, such as deep convolution networks, deep residual networks, recurrent neural networks, reinforcement learning, variational autoencoders, and others.

/pdf/review-of-deep-learning-algorithms-and-architectures-41h40bgta6.pdf

Review of Deep Learning Algorithms and Architectures

A mobile robot guest for interacting with a human resident performs a room-traversing search procedure prior to interacting with the resident, and may verbally query whether the resident being sought is present. Upon finding the resident, the mobile robot may facilitate a teleconferencing session with a remote third party, or interact with the resident in a number of ways. For example, the robot may carry on a dialogue with the resident, reinforce compliance with medication or other schedules, etc. In addition, the robot incorporates safety features for preventing collisions with the resident; and the robot may audibly announce and/or visibly indicate its presence in order to avoid becoming a dangerous obstacle. Furthermore, the mobile robot behaves in accordance with an integral privacy policy, such that any sensor recording or transmission must be approved by the resident.

Companion robot for personal interaction

Audio-visual speech recognition (AVSR) system is thought to be one of the most promising solutions for reliable speech recognition, particularly when the audio is corrupted by noise. However, cautious selection of sensory features is crucial for attaining high recognition performance. In the machine-learning community, deep learning approaches have recently attracted increasing attention because deep neural networks can effectively extract robust latent features that enable various recognition algorithms to demonstrate revolutionary generalization capabilities under diverse application conditions. This study introduces a connectionist-hidden Markov model (HMM) system for noise-robust AVSR. First, a deep denoising autoencoder is utilized for acquiring noise-robust audio features. By preparing the training data for the network with pairs of consecutive multiple steps of deteriorated audio features and the corresponding clean features, the network is trained to output denoised audio features from the corresponding features deteriorated by noise. Second, a convolutional neural network (CNN) is utilized to extract visual features from raw mouth area images. By preparing the training data for the CNN as pairs of raw images and the corresponding phoneme label outputs, the network is trained to predict phoneme labels from the corresponding mouth area input images. Finally, a multi-stream HMM (MSHMM) is applied for integrating the acquired audio and visual HMMs independently trained with the respective features. By comparing the cases when normal and denoised mel-frequency cepstral coefficients (MFCCs) are utilized as audio features to the HMM, our unimodal isolated word recognition results demonstrate that approximately 65 % word recognition rate gain is attained with denoised MFCCs under 10 dB signal-to-noise-ratio (SNR) for the audio signal input. Moreover, our multimodal isolated word recognition results utilizing MSHMM with denoised MFCCs and acquired visual features demonstrate that an additional word recognition rate gain is attained for the SNR conditions below 10 dB.

/pdf/audio-visual-speech-recognition-using-deep-learning-2nrljpc31c.pdf

Audio-visual speech recognition using deep learning

Multimodal integration learning of robot behavior using deep neural networks

2006 Special issue: Dynamic and interactive generation of object handling behaviors by a small humanoid robot using a dynamic neural network model

In recent automatic speech recognition studies, deep learning architecture applications for acoustic modeling have eclipsed conventional sound features such as Mel-frequency cepstral coefficients. However, for visual speech recognition (VSR) studies, handcrafted visual feature extraction mechanisms are still widely utilized. In this paper, we propose to apply a convolutional neural network (CNN) as a visual feature extraction mechanism for VSR. By training a CNN with images of a speaker’s mouth area in combination with phoneme labels, the CNN acquires multiple convolutional filters, used to extract visual features essential for recognizing phonemes. Further, by modeling the temporal dependencies of the generated phoneme label sequences, a hidden Markov model in our proposed system recognizes multiple isolated words. Our proposed system is evaluated on an audio-visual speech dataset comprising 300 Japanese words with six different speakers. The evaluation results of our isolated word recognition experiment demonstrate that the visual features acquired by the CNN significantly outperform those acquired by conventional dimensionality compression approaches, including principal component analysis.

Lipreading using convolutional neural network

Recognizing grasped objects with tactile sensors is beneficial in many situations, as other sensor information like vision is not always reliable. In this paper, we aim for multimodal object recognition by power grasping of objects with an unknown orientation and position relation to the hand. Few robots have the necessary tactile sensors to reliably recognize objects: in this study the multifingered hand of TWENDY-ONE is used, which has distributed skin sensors covering most of the hand, 6 axis F/T sensors in each fingertip, and provides information about the joint angles. Moreover, the hand is compliant. When using tactile sensors, it is not clear what kinds of features are useful for object recognition. Recently, deep learning has shown promising results. Nevertheless, deep learning has rarely been used in robotics and to our best knowledge never for tactile sensing, probably because it is difficult to gather many samples with tactile sensors. Our results show a clear improvement when using a denoising autoencoder with dropout compared to traditional neural networks. Nevertheless, a higher number of layers did not prove to be beneficial.

Kuniaki Noda

Papers

Audio-visual speech recognition using deep learning

Multimodal integration learning of robot behavior using deep neural networks

2006 Special issue: Dynamic and interactive generation of object handling behaviors by a small humanoid robot using a dynamic neural network model

Lipreading using convolutional neural network

Tactile object recognition using deep learning and dropout