Customer-obsessed science
-
January 13, 2023Using lists of rare or out-of-vocabulary words to bias connectionist temporal classification models enables personalization. -
January 11, 2023Private aggregation of teacher ensembles (PATE) leads to word error rate reductions of more than 26% relative to standard differential-privacy techniques.
-
January 06, 2023Quantization with self-adjustable centroids, contrastive predictive coding for transfer learning, teacher ensembles for differential privacy, and more — Amazon’s speech research features a battery of cutting-edge machine learning techniques.
-
-
February 4 - 10, 2023
-
February 7 - 14, 2023
-
February 27 - March 3, 2023
-
-
Tennessee State UniversityJanuary 12, 2023The collaboration, housed in the College of Engineering, includes funds for faculty research projects, with an initial focus on AI, robotics, and operations research.
-
January 09, 2023Mehrdad Mahdjoubi, founder and CEO of Alexa Fund portfolio company, explains ‘no compromise’ approach to saving resources without sacrificing user experience.
-
January 05, 2023How an AWS customer uses Lookout for Vision to build custom computer vision models to automate quality inspection and detect defects.
-
December 30, 2022From the remarkable story of Josh Miele and his passion for improving accessibility, to the challenges overcome by scientists and engineers getting Alexa to work in space, these five stories touched both emotional and intellectual chords.
-
Who moderates on Twitch and what do they do? Quantifying practices in community moderation on TwitchGROUP 20232023Volunteer moderators are an increasingly essential component of effective community management across a range of services, such as Facebook, Reddit, Discord, YouTube, and Twitch. Prior work has investigated how users of these services become moderators, their attitudes towards community moderation, and the work that they perform, largely through interviews with community moderators and managers. In this
-
Data augmentation is a necessity to enhance data efficiency in deep learning. For vision-language pre-training, data is only augmented either for images or for text in previous works. In this paper, we present MixGen: a joint data augmentation for vision-language representation learning to further improve data efficiency. It generates new imagetext pairs with semantic relationships preserved by interpolating
-
2023Semantic matching is an important component of a product search pipeline. Its goal is to capture the semantic intent of the search query as opposed to the syntactic matching performed by a lexical matching system. A semantic matching model captures relationships like synonyms, and also captures common behavioral patterns to retrieve relevant results by generalizing from purchase data. Semantic matching
-
Conversational agents like chat bots and voice assistants are trained to understand and respond to user intents. On encountering an utterance with an intent different from the ones they have been trained on, these agents are expected to classify the intent as “unknown” or “out of domain”. This problem is known as out of domain (OOD) intent detection. Podolskiy et al. (2021), showed that Mahalanobis distance
-
Video quality assessment (VQA) has sparked a lot of interest in the computer vision community, as it plays a critical role in services that provide customers with high quality video content. Due to the lack of high quality reference videos and the difficulties in collecting subjective evaluations, assessing video quality is a challenging and still unsolved problem. Moreover, most of the public research
-
December 23, 2022Program focuses on diversifying tech-industry talent.
-
December 21, 2022Fifth challenge adds new elements and features four new competitors for the $1 million research grant.
-
Courtesy of Yong Jae LeeDecember 16, 2022University of Wisconsin-Madison associate professor and ARA recipient has authored a series of pioneering papers on real-time object instance segmentation.
Working at Amazon
View allMeet the people driving the innovation essential to being the world’s most customer-centric company.
View all
-
Courtesy of InterspeechDecember 20, 2022Ariadna Sanchez, a scientist who works in polyglot text to speech, draws on her musical background to help find novel solutions.
-
Courtesy of Jiao SunDecember 13, 2022Amazon Machine Learning Fellow Jiao Sun works on strategies to control text generation.
-
November 09, 2022At Amazon, he develops machine learning models to help keeping Amazon stores safe and trustworthy for customers and selling partners.
