Amazon Scholars and Amazon Visiting Academics use AWS to accelerate research
The Amazon Web Services (AWS) Cloud Credit for Research Program launched a new opportunity for Amazon Scholars and Amazon Visiting Academics to apply for AWS Promotional Credit to accelerate innovation through cloud technology.
Amazon Scholars are world-class academics and Amazon Visiting Academics are pre- to newly-tenured academics, selected to tackle real-world technical challenges as they continue to teach and conduct research at their universities. AWS Promotional Credit is offered through this initiative is to support the awardees’ impactful university research that’s distinct from their work at Amazon.
Amazon Scholars and Amazon Visiting Academics come from all around the globe to innovate in research areas including, using cloud services like machine learning (ML), computer vision, data science, natural language processing (NLP), robotics, economics, optimization, and quantum computing. Harnessing their expertise, Amazon Scholars and Amazon Visiting Academics advise Amazon leaders on strategic plans, dive deep to solve a specific technical problem in an organization’s roadmap, and advise junior researchers on methods. For example, Amazon Scholar Julia Hirschberg is integral to natural turn-taking, a to-be-launched feature that helps Alexa interact more naturally with people without needing to repeat the wake word “Alexa,” and Amazon Visiting Academic James Whitfield is educating AWS customers on quantum computing and how to leverage quantum technologies to address needs in fields such as data security and materials development.
After a detailed review of proposals, AWS selected 34 researchers to receive AWS Promotional Credit to support their university projects. Check out a sample of AWS Promotional Credit recipients and their proposed initiatives:
- Mohsen Bayati, associate professor at the Stanford Graduate School of Business, is collaborating with the Stanford Cancer Center to create clinically relevant covariates that can predict the recurrence of cancer and help personalize treatment plans. This ML model will be able to synthesize large volumes of patient data and has the potential to act as a decision-support assistant for oncologists.
- Brian Kulis, associate professor in department of electrical and computer engineering and the department of computer science at Boston University, is leveraging deep learning to make automatic music generation (using technology to generate or modify melodies and rhythms) easier and more effective for artists and music content providers. This development provides content providers a way to create certain kinds of music (e.g. mood or background music) without the potentially expensive licensing fees.
- Heng Ji, professor of computer science at the University of Illinois at Urbana-Champaign, is proposing a novel graphical neural network model that processes multilingual multimedia news to detect artificial intelligence (AI)-generated misinformation. Through a new “information surgeon” model that harnesses state-of-the-art extraction techniques, fake news will be identified and regenerated to reflect the truth.
Amazon Visiting Academics
- Emilio Ferrara, associate professor of communication, computer science, preventive medicine at the University of Southern California, is engineering a multimodal ML model that can detect misinformation, disinformation, and influence online. The model is set to analyze real-time COVID-19 online discourse for anti-science campaigns, as well as online political influence campaigns from foreign adversaries.
- Marco Giometto, assistant professor of civil engineering and engineering mechanics at Columbia University, is developing a 3D physics-informed ML model that reconstructs the 3D structure of atmospheric air pollutants from 2D satellite images thereof. The model would provide information such as where the maximum pollutant concentration is located, and in doing so help inform effective response strategies (e.g. evacuation pathways following wildfires or release of toxic substances).
- Nanyun (Violet) Peng, assistant professor of computer science at the University of California, Los Angeles (UCLA), is advancing narrative language processing by training machines to understand the different components of a narrative, including characters, events and their relations, and social and temporal commonsense. This helps AI systems better understand emotions, everyday actions, factual knowledge, and commonsense reasoning to improve real-world applications such as personal assistant systems.
Learn more about the AWS Cloud Credit for Research Program, Amazon Scholars, and Amazon Visiting Academics. Accomplished academics who are interested in learning more about how their research may match the challenges, opportunities, and scale of Amazon are encouraged to reach out to email@example.com for more information.