PERSONALIZA TION OF END-TO-END SPEECH RECOGNITION ON MOBILE DEVICES FOR NAMED ENTITIES Khe Chai Sim, Franc oise Beaufays, Arnaud Benard, Dhruv Guliani, Andreas Kabel, Nikhil Khare, T amar Lucassen, Petr Zadrazil, Harry Zhang, Leif Johnson, Giovanni Motta, Lillian Zhou Google, USA ABSTRACT We study the effectiveness of several techniques to personalize end-to-end speech models and improve the recognition of proper names relevant to the user. These techniques differ in the amounts of user effort required to provide supervision, and are evaluated on how they impact speech recognition performance. We propose using keyword-dependent precision and recall metrics to measure vocabulary acquisition performance. We evaluate the algorithms on a dataset that we designed to contain names of persons that are difficult to recognize. Therefore, the baseline recall rate for proper names in this dataset is very low: 2.4%. A data synthesis approach we developed brings it to 48.6%, with no need for speech input from the user. With speech input, if the user corrects only the names, the name recall rate improves to 64.4%. If the user corrects all the recognition errors, we achieve the best recall of 73.5%. To eliminate the need to upload user data and store personalized models on a server, we focus on performing the entire personalization workflow on a mobile device.

Sleep apnea is a common respiratory disorder characterized by breathing pauses during the night. Consequences of untreated sleep apnea can be severe. Still, many people remain undiagnosed due to shortages of hospital beds and trained sleep technicians. To assist in the diagnosis process, automated detection methods are being developed. Recent works have demonstrated that deep learning models can extract useful information from raw respiratory data and that such models can be used as a robust sleep apnea detector. However, trained sleep technicians take into account multiple sensor signals when annotating sleep recordings instead of relying on a single respiratory estimate. To improve the predictive performance and reliability of the models, early and late sensor fusion methods are explored in this work. In addition, a novel late sensor fusion method is proposed which uses backward shortcut connections to improve the learning of the first stages of the models. The performance of these fusion methods is analyzed using CNN as well as LSTM deep learning base-models. The results demonstrate a significant and consistent improvement in predictive performance over the single sensor methods and over the other explored sensor fusion methods, by using the proposed sensor fusion method with backward shortcut connections.

We propose a new contextual-compositional neural network layer that handles out-of-vocabulary (OOV) words in natural language processing (NLP) tagging tasks. This layer consists of a model that attends to both the character sequence and the context in which the OOV words appear. We show that our model learns to generate task-specific \textit{and} sentence-dependent OOV word representations without the need for pre-training on an embedding table, unlike previous attempts. We insert our layer in the state-of-the-art tagging model of \citet{plank2016multilingual} and thoroughly evaluate its contribution on 23 different languages on the task of jointly tagging part-of-speech and morphosyntactic attributes. Our OOV handling method successfully improves performances of this model on every language but one to achieve a new state-of-the-art on the Universal Dependencies Dataset 1.4.

Deep neural networks are a promising approach towards multi-task learning because of their capability to leverage knowledge across domains and learn general purpose representations. Nevertheless, they can fail to live up to these promises as tasks often compete for a model's limited resources, potentially leading to lower overall performance. In this work we tackle the issue of interfering tasks through a comprehensive analysis of their training, derived from looking at the interaction between gradients within their shared parameters. Our empirical results show that well-performing models have low variance in the angles between task gradients and that popular regularization methods implicitly reduce this measure. Based on this observation, we propose a novel gradient regularization term that minimizes task interference by enforcing near orthogonal gradients. Updating the shared parameters using this property encourages task specific decoders to optimize different parts of the feature extractor, thus reducing competition. We evaluate our method with classification and regression tasks on the multiDigitMNIST, NYUv2 and SUN RGB-D datasets where we obtain competitive results.

This paper studies the fundamental problem of learning deep generative models that consist of one or more layers of latent variables organized in top-down architectures. Learning such a generative model requires inferring the latent variables for each training example based on the posterior distribution of these latent variables. The inference typically requires Markov chain Monte Caro (MCMC) that can be time consuming. In this paper, we propose to use short run inference dynamics guided by the log-posterior, such as finite-step gradient descent algorithm initialized from the prior distribution of the latent variables, as an approximate sampler of the posterior distribution, where the step size of the gradient descent dynamics is optimized by minimizing the Kullback-Leibler divergence between the distribution produced by the short run inference dynamics and the posterior distribution. Our experiments show that the proposed method outperforms variational auto-encoder (VAE) in terms of reconstruction error and synthesis quality. The advantage of the proposed method is that it is natural and automatic, even for models with multiple layers of latent variables.

Intelligent transportation systems (ITSs) are envisioned to be crucial for smart cities, which aims at improving traffic flow to improve the life quality of urban residents and reducing congestion to improve the efficiency of commuting. However, several challenges need to be resolved before such systems can be deployed, for example, conventional solutions for Markov decision process (MDP) and single-agent Reinforcement Learning (RL) algorithms suffer from poor scalability, and multi-agent systems suffer from poor communication and coordination. In this paper, we explore the potential of mutual information sharing, or in other words, spatial influence based communication, to optimize traffic light control policy. First, we mathematically analyze the transportation system. We conclude that the transportation system does not have stationary Nash Equilibrium, thereby reinforcement learning algorithms offer suitable solutions. Secondly, we describe how to build a multi-agent Deep Deterministic Policy Gradient (DDPG) system with spatial influence and social group utility incorporated. Then we utilize the grid topology road network to empirically demonstrate the scalability of the new system. We demonstrate three types of directed communications to show the effect of directions of social influence on the entire network utility and individual utility. Lastly, we define "selfish index" and analyze the effect of it on total group utility.

With the turmoil in cybersecurity and the mind-blowing advances in AI, it is only natural that cybersecurity practitioners consider further employing learning techniques to help secure their organizations and improve the efficiency of their security operation centers. But with great fears come great opportunities for both the good and the evil, and a myriad of bad deals. This paper discusses ten issues in cybersecurity that hopefully will make it easier for practitioners to ask detailed questions about what they want from an AI system in their cybersecurity operations. We draw on the state of the art to provide factual arguments for a discussion on well-established AI in cybersecurity issues, including the current scope of AI and its application to cybersecurity, the impact of privacy concerns on the cybersecurity data that can be collected and shared externally to the organization, how an AI decision can be explained to the person running the operations center, and the implications of the adversarial nature of cybersecurity in the learning techniques. We then discuss the use of AI by attackers on a level playing field including several issues in an AI battlefield, and an AI perspective on the old cat-and-mouse game including how the adversary may assess your AI power.

Hopefully, this app helps you to accelerate your next-generation AI or ML initiative leveraging Splunk's Data-To-Everything-Platform and your favourite frameworks or open source libraries. You find many Juypter notebook examples and a predefined workflow that should help you to get started easily! Even if the desired set of ML libraries is not there yet, you can easily extend the app with your custom MLTK Container. Rebuild the existing MLTK Container images or build your own custom images with the open-source repository on GitHub. Most recently, we finally released the latest version of Deep Learning Toolkit 3.0 which is compatible with Splunk 8.0 and Machine Learning Toolkit 5.0 based on Python 3.

More than 13,000 artificial intelligence mavens flocked to Vancouver this week for the world's leading academic AI conference, NeurIPS. The venue included a maze of colorful corporate booths aiming to lure recruits for projects like software that plays doctor. Google handed out free luggage scales and socks depicting the colorful bikes employees ride on its campus while IBM offered hats emblazoned with "I A ." Tuesday night, Google and Uber hosted well-lubricated, over-subscribed parties. At a bleary 8:30 the next morning, one of Google's top researchers gave a keynote with a sobering message about AI's future. Blaise Aguera y Arcas praised the revolutionary technique known as deep learning that has seen teams like his get phones to recognize faces and voices.

Rome wasn't built in a day. It has taken years for computers to exhibit the level of intelligence they do today and be able to produce text that sounds and read human-like. It's time to appreciate this revolutionary journey. In 2007, The first step was taken by Robbie Allen, who was a veteran engineer at Cisco. He created an online college basketball website that automatically published game reviews, real-time updates, recaps, and incidents of injury.