Search results

Retail websites may provide customers with a localized user experience by allowing them to use a secondary language of preference. Automatic translation of user search queries is a crucial component of this experience. Several domain-adapted SMT systems for search query translation were trained, including language pairs for which smaller-than desired parallel resources were available, such as Polish-German

We consider the teacher-student framework for knowledge transfer, where the goal is to improve learning of a “student” neural network, given a “teacher” neural network pretrained on the same or a similar task. The majority of existing approaches for distilling knowledge from a teacher network to a student network rely on matching either activations or handcrafted features from the teacher network. Instead

Continual learning aims to enable machine learning models to learn a general solution space for past and future tasks in a sequential manner. Conventional models tend to forget the knowledge of previous tasks while learning a new task, a phenomenon known as catastrophic forgetting. When using Bayesian models in continual learning, knowledge from previous tasks can be retained in two ways: (i) posterior

In complex transfer learning scenarios new tasks might not be tightly linked to previous tasks. Approaches that transfer information contained only in the final parameters of a source model will therefore struggle. Instead, transfer learning at a higher level of abstraction is needed. We propose Leap, a framework that achieves this by transferring knowledge across learning processes. We associate each task

We tackle the problem of optimizing a blackbox objective function defined over a highly structured input space. This problem is ubiquitous in machine learning. Inferring the structure of a neural network or the Automatic Statistician (AS), where the kernel combination for a Gaussian process is optimized, are two of many possible examples. We use the AS as a case study to describe our approach, that can

Training large neural networks requires distributing learning across multiple workers, where the cost of communicating gradients can be a significant bottleneck. SIGNSGD alleviates this problem by transmitting just the sign of each minibatch stochastic gradient. We prove that it can get the best of both worlds: compressed gradients and SGD-level convergence rate. The relative `1/`2 geometry of gradients

We revisit the problem of linear regression under a differential privacy constraint. By consolidating existing pieces in the literature, we clarify the correct dependence of the feature, label and coefficient domains in the optimization error and estimation error, hence revealing the delicate price of differential privacy in statistical estimation and statistical learning. Moreover, we propose simple modifications

Knowledge-based question answering relies on the availability of facts, the majority of which cannot be found in structured sources (e.g. Wikipedia info-boxes, Wikidata). One of the major components of extracting facts from unstructured text is Relation Extraction (RE). In this paper we propose a novel method for creating distant (weak) supervision labels for training a large-scale RE system. We also provide

We propose a framework that harnesses visual cues in an unsupervised manner to learn the co-occurrence distribution of items in real-world images for complementary recommendation. Our model learns a non-linear transformation between the two manifolds of source and target item categories (e.g., tops and bottoms in outfits). Given a large dataset of images containing instances of co-occurring items, we train

Recent work has made significant progress in improving spatial resolution for pixelwise labeling with Fully Convolutional Network (FCN) framework by employing Dilated/Atrous convolution, utilizing multi-scale features and refining boundaries. In this paper, we explore the impact of global contextual information in semantic segmentation by introducing the Context Encoding Module, which captures the semantic

Deep learning models have become state of the art for natural language processing (NLP) tasks, however deploying these models in production system poses significant memory constraints. Existing compression methods are either lossy or introduce significant latency. We propose a compression method that leverages low rank matrix factorization during training, to compress the word embedding layer which represents

We describe an intelligent context-aware conversational system that incorporates screen context information to service multimodal user requests. Screen content is used for disambiguation of utterances that refer to screen objects and for enabling the user to act upon screen objects using voice commands. We propose a deep learning architecture that jointly models the user utterance and the screen and incorporates

Pitch detection is a fundamental problem in speech processing as F0 is used in a large number of applications. Recent papers have proposed deep learning for robust pitch tracking. In this letter, we consider voicing detection as a classification problem and F0 contour estimation as a regression problem. For both tasks, acoustic features from multiple domains and traditional machine learning methods are

Accurate on-device wake word detection is crucial to products with far-field voice control such as the Amazon Echo. It is quite challenging to build a wake word system with both low False Reject Rate (FRR) and low False Alarm Rate (FAR) in real scenarios where there are various types of background speech, music or noise, especially when computational resources on the device is limited. In this paper, we

We propose a novel information-theoretic approach for evaluating microphone arrays that relies on the array physics and geometry rather than the underlying beamforming algorithm. The analogy between Multiple-Input-Multiple-Output (MIMO) wireless communication channel and the acoustic channel of microphone arrays is exploited to define information measures of microphone arrays, which provide upper bounds

We propose a simple recurrent model for detecting rare sound events, when the time boundaries of events are available for training. Our model optimizes the combination of an utterancelevel loss, which classifies whether an event occurs in an utterance, and a frame-level loss, which classifies whether each frame corresponds to the event when it does occur. The two losses make use of a shared vectorial representation

This paper proposes an efficient real-time multirate fast transient-sound detection algorithm on the basis of emerging microphone array configuration intended for multimedia signal processing application systems such as digital smart home. The proposed detection algorithm first extracts the dynamics and periodicity features, then trains the model parameters of these features on Amazon machine learning platform

This paper proposes a Region-based Convolutional Recurrent Neural Network (R-CRNN) for audio event detection (AED). The proposed network is inspired by Faster-RCNN [1], a wellknown region-based convolutional network framework for visual object detection. Different from the original Faster-RCNN, a recurrent layer is added on top of the convolutional network to capture the long-term temporal context from

Multimedia streaming services over spoken dialog systems have become ubiquitous. User-entity affinity modeling is critical for the system to understand and disambiguate user intents and personalize user experiences. However, fully voice-based interaction demands quantification of novel behavioral cues to determine user affinities. In this work, we propose using play duration cues to learn a matrix factorization

Recurrent Neural Networks (RNN) have recently proved to be effective in acoustic modeling for TTS. Various techniques such as the Maximum Likelihood Parameter Generation (MLPG) algorithm have been naturally inherited from the HMM-based speech synthesis framework. This paper investigates in which situations parameter generation and variance restoration approaches help for RNN-based TTS. We explore how their