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.

Universal Sentence Encoder for English

Abstract: We present easy-to-use TensorFlow Hub sentence embedding models having good task transfer performance. Model variants allow for trade-offs between accuracy and compute resources. We report the relationship between model complexity, resources, and transfer performance. Comparisons are made with baselines without transfer learning and to baselines that incorporate word-level transfer. Transfer learning using sentence-level embeddings is shown to outperform models without transfer learning and often those that use only word-level transfer. We show good transfer task performance with minimal training data and obtain encouraging results on word embedding association tests (WEAT) of model bias.

Networked personal contact manager

Abstract: A network-computer-based personal contact manager system is disclosed wherein users of networked clients maintain and update a set of user information which is stored in a relational database on a networked server. The personal contact manager system allows each user to specify on an individual basis which of their contacts are permitted to access respective datums of their user information. In some cases, and assuming permission is granted, the system will issue notifications (e.g., by e-mail) to a user's contacts when the user changes his information or when a preset event, such as a birthday, as defined by the user, is to occur. The system also allows users to find contacts based on common group affiliations and notifies users when there are coincidences in their data (e.g., travel plans, astrological compatibility). The personal contact manager system supports the retrieval of information on the contacts of contacts, assuming such as permission has been granted by the contacts and their contacts, and can also be used to synchronize the server database with a PIM database of the user and any contacts of the user whoe have the appropriate permissions.

A reconfigurable fabric for accelerating large-scale datacenter services

Abstract: Datacenter workloads demand high computational capabilities, flexibility, power efficiency, and low cost It is challenging to improve all of these factors simultaneously To advance datacenter capabilities beyond what commodity server designs can provide, we designed and built a composable, reconfigurable hardware fabric based on field programmable gate arrays (FPGA) Each server in the fabric contains one FPGA, and all FPGAs within a 48-server rack are interconnected over a low-latency, high-bandwidth networkWe describe a medium-scale deployment of this fabric on a bed of 1632 servers, and measure its effectiveness in accelerating the ranking component of the Bing web search engine We describe the requirements and architecture of the system, detail the critical engineering challenges and solutions needed to make the system robust in the presence of failures, and measure the performance, power, and resilience of the system Under high load, the large-scale reconfigurable fabric improves the ranking throughput of each server by 95% at a desirable latency distribution or reduces tail latency by 29% at a fixed throughput In other words, the reconfigurable fabric enables the same throughput using only half the number of servers

ResNeSt: Split-Attention Networks

Abstract: It is well known that featuremap attention and multi-path representation are important for visual recognition. In this paper, we present a modularized architecture, which applies the channel-wise attention on different network branches to leverage their success in capturing cross-feature interactions and learning diverse representations. Our design results in a simple and unified computation block, which can be parameterized using only a few variables. Our model, named ResNeSt, outperforms EfficientNet in accuracy and latency trade-off on image classification. In addition, ResNeSt has achieved superior transfer learning results on several public benchmarks serving as the backbone, and has been adopted by the winning entries of COCO-LVIS challenge. The source code for complete system and pretrained models are publicly available.