Amazon.com

About: Amazon.com is a company organization based out in Seattle, Washington, United States. It is known for research contribution in the topics: Computer science & Service (business). The organization has 13363 authors who have published 17317 publications receiving 266589 citations.

Jasper: An End-to-End Convolutional Neural Acoustic Model.

Abstract: In this paper, we report state-of-the-art results on LibriSpeech among end-to-end speech recognition models without any external training data. Our model, Jasper, uses only 1D convolutions, batch normalization, ReLU, dropout, and residual connections. To improve training, we further introduce a new layer-wise optimizer called NovoGrad. Through experiments, we demonstrate that the proposed deep architecture performs as well or better than more complex choices. Our deepest Jasper variant uses 54 convolutional layers. With this architecture, we achieve 2.95% WER using a beam-search decoder with an external neural language model and 3.86% WER with a greedy decoder on LibriSpeech test-clean. We also report competitive results on the Wall Street Journal and the Hub5'00 conversational evaluation datasets.

Performing automatically authorized programmatic transactions

Abstract: Techniques are described for facilitating interactions between computing systems, such as by performing transactions between parties that are automatically authorized via a third-party transaction authorization system. In some situations, the transactions are programmatic transactions involving the use of fee-based Web services by executing application programs, with the transaction authorization system authorizing and/or providing payments in accordance with private authorization instructions previously specified by the parties. The authorization instructions may include predefined instruction rule sets that regulate conditions under which a potential transaction can be authorized, with the instruction rule sets each referenced by an associated reference token. After one or more of the parties to a potential transaction supply reference tokens for the parties, the transaction authorization system can determine whether to authorize the transaction based on whether the instruction rule sets associated with the reference tokens are compatible or otherwise satisfied.

A data-centric approach to insider attack detection in database systems

Abstract: The insider threat against database management systems is a dangerous security problem. Authorized users may abuse legitimate privileges to masquerade as other users or to maliciously harvest data. We propose a new direction to address this problem. We model users' access patterns by profiling the data points that users access, in contrast to analyzing the query expressions in prior approaches. Our data-centric approach is based on the key observation that query syntax alone is a poor discriminator of user intent, which is much better rendered by what is accessed. We present a feature-extraction method to model users' access patterns. Statistical learning algorithms are trained and tested using data from a real Graduate Admission database. Experimental results indicate that the technique is very effective, accurate, and is promising in complementing existing database security solutions. Practical performance issues are also addressed.

Multi-style Generative Network for Real-Time Transfer

Abstract: Despite the rapid progress in style transfer, existing approaches using feed-forward generative network for multi-style or arbitrary-style transfer are usually compromised of image quality and model flexibility. We find it is fundamentally difficult to achieve comprehensive style modeling using 1-dimensional style embedding. Motivated by this, we introduce CoMatch Layer that learns to match the second order feature statistics with the target styles. With the CoMatch Layer, we build a Multi-style Generative Network (MSG-Net), which achieves real-time performance. In addition, we employ an specific strategy of upsampled convolution which avoids checkerboard artifacts caused by fractionally-strided convolution. Our method has achieved superior image quality comparing to state-of-the-art approaches. The proposed MSG-Net as a general approach for real-time style transfer is compatible with most existing techniques including content-style interpolation, color-preserving, spatial control and brush stroke size control. MSG-Net is the first to achieve real-time brush-size control in a purely feed-forward manner for style transfer. Our implementations and pre-trained models for Torch, PyTorch and MXNet frameworks will be publicly available (Links can be found at http://hangzhang.org/).

Street-view change detection with deconvolutional networks

Abstract: We propose a system for performing structural change detection in street-view videos captured by a vehicle-mounted monocular camera over time. Our approach is motivated by the need for more frequent and efficient updates in the large-scale maps used in autonomous vehicle navigation. Our method chains a multi-sensor fusion SLAM and fast dense 3D reconstruction pipeline, which provide coarsely registered image pairs to a deep Deconvolutional Network (DN) for pixel-wise change detection. We investigate two DN architectures for change detection, the first one is based on the idea of stacking contraction and expansion blocks while the second one is based on the idea of Fully Convolutional Networks. To train and evaluate our networks we introduce a new urban change detection dataset which is an order of magnitude larger than existing datasets and contains challenging changes due to seasonal and lighting variations. Our method outperforms existing literature on this dataset, which we make available to the community, and an existing panoramic change detection dataset, demonstrating its wide applicability.