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.

Request routing in a networked environment

Abstract: A system, methods, and interfaces for managing request routing functionality associated with resource requests for one or more resources associated with a content provider. The request routing functionality can correspond to the processing of domain name service (“DNS”) requests for resources by computing devices and the resolution of the DNS requests by the identification of a network address of a computing device that will provide the requested resources. Unlike traditional CDN service provider implementation, the processing of resource requests by the service provider is separate from the delivery of the content by the content provider (or on behalf of the content provider).

Countermeasures for threats to an uncrewed autonomous vehicle

Abstract: Uncrewed autonomous vehicles (“UAVs”) may navigate from one location to another location. Described herein are systems, devices, and methods providing countermeasures for threats that may compromise the UAVs. A plurality of UAVs may establish a mesh network to distribute information to one another. A first UAV may receive external data from a second UAV using the mesh network. The external data may be used to confirm or cross-check data such as location, heading, altitude, and so forth. Disagreement between data generated by the first UAV with external data from the second UAV may result in the determination that the first UAV is compromised. Remedial actions may be taken, such as the first UAV may be directed to a safe location to land or park, may receive commands from another UAV, and so forth.

SUPERB: Speech processing Universal PERformance Benchmark

Abstract: Self-supervised learning (SSL) has proven vital for advancing research in natural language processing (NLP) and computer vision (CV). The paradigm pretrains a shared model on large volumes of unlabeled data and achieves state-of-the-art (SOTA) for various tasks with minimal adaptation. However, the speech processing community lacks a similar setup to systematically explore the paradigm. To bridge this gap, we introduce Speech processing Universal PERformance Benchmark (SUPERB). SUPERB is a leaderboard to benchmark the performance of a shared model across a wide range of speech processing tasks with minimal architecture changes and labeled data. Among multiple usages of the shared model, we especially focus on extracting the representation learned from SSL due to its preferable re-usability. We present a simple framework to solve SUPERB tasks by learning task-specialized lightweight prediction heads on top of the frozen shared model. Our results demonstrate that the framework is promising as SSL representations show competitive generalizability and accessibility across SUPERB tasks. We release SUPERB as a challenge with a leaderboard and a benchmark toolkit to fuel the research in representation learning and general speech processing.