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: Service (business) & Service provider. The organization has 13363 authors who have published 17317 publications receiving 266589 citations.

Gestures and touches on force-sensitive input devices

Abstract: A force-sensitive touch sensor detects location and force of touches applied to the sensor. Movement of an object touching the force-sensitive touch sensor correlates to movement of a pointer on a display device. Varying levels of force applied to the force-sensitive touch sensor are interpreted as different commands. Objects displayed on the display device can be manipulated by a combination of gestures across a surface of the force-sensitive touch sensor and changes in force applied to the force-sensitive touch sensor.

Topic Modeling with Wasserstein Autoencoders

Abstract: We propose a novel neural topic model in the Wasserstein autoencoders (WAE) framework. Unlike existing variational autoencoder based models, we directly enforce Dirichlet prior on the latent document-topic vectors. We exploit the structure of the latent space and apply a suitable kernel in minimizing the Maximum Mean Discrepancy (MMD) to perform distribution matching. We discover that MMD performs much better than the Generative Adversarial Network (GAN) in matching high dimensional Dirichlet distribution. We further discover that incorporating randomness in the encoder output during training leads to significantly more coherent topics. To measure the diversity of the produced topics, we propose a simple topic uniqueness metric. Together with the widely used coherence measure NPMI, we offer a more wholistic evaluation of topic quality. Experiments on several real datasets show that our model produces significantly better topics than existing topic models.

Multilingual Sequence-to-Sequence Speech Recognition: Architecture, Transfer Learning, and Language Modeling

Abstract: Sequence-to-sequence (seq2seq) approach for low-resource ASR is a relatively new direction in speech research. The approach benefits by performing model training without using lexicon and alignments. However, this poses a new problem of requiring more data compared to conventional DNN-HMM systems. In this work, we attempt to use data from 10 BABEL languages to build a multilingual seq2seq model as a prior model, and then port them towards 4 other BABEL languages using transfer learning approach. We also explore different architectures for improving the prior multilingual seq2seq model. The paper also discusses the effect of integrating a recurrent neural network language model (RNNLM) with a seq2seq model during decoding. Experimental results show that the transfer learning approach from the multilingual model shows substantial gains over monolingual models across all 4 BABEL languages. Incorporating an RNNLM also brings significant improvements in terms of %WER, and achieves recognition performance comparable to the models trained with twice more training data.

Time-based warehouse movement maps

Abstract: A computer software facility for generating and implementing a temporal map depicting a relationship between locations in a warehouse or distribution center is described. The facility tracks the movement in the warehouse, including the time it takes for mobile elements to move between a pair of identifiable locations. Based on the tracking, the facility produces movement time records for multiple location pairs in the warehouse. The facility then generates a map of the warehouse. Where information for specific pairs of identifiable locations is not available, the facility may derive movement time information based on actual movement times for related location pairs. The generated map can be used as part of a computer application for tasks such as scheduling the picking of items, evaluating employee performance, organizing the storage of items in the warehouse, and other uses.