Faculty: Alexander I. Rudnicky
Alexa Fellow: Ran Zhao
Alexa Fellow Bio: Ran Zhao received his B.S. in Computer Science from University of Illinois, Urbana-Champaign working with Prof. Dan Roth in Natural Language Processing and a M.S. from Yale University. His thesis focuses on extending task-oriented dialogue systems with social capabilities to better reflect the characteristics of human-human interaction. His work has contributed significantly to theoretical research in social science and application of human-computer interaction. He is an author on 11 papers and received the Best Student Paper at the 16th International Conference on Intelligent Virtual Agents. From 2017 to present, he has been the Amazon Alexa Fellow at Carnegie Mellon.
Course description: Amazon Alexa is being used in courses at Carnegie Mellon to introduce students to basic ideas in spoken language interaction. Students learn how to structure conversation and how to manage task-oriented language understanding. Alexa Skills Kit allows students to quickly build systems by providing base speech recognition and speech synthesis capabilities, allowing them to focus on interaction-level problems.
Faculty: Alexander Wong, Igor Ivkovic, Shelley Wang
Alexa Fellow: TBA
Alexa Fellow Bio: TBA
Course description: Fundamentals of Computational Intelligence (ECE 457B): Introduces novel approaches for computational intelligence based techniques including: knowledge based reasoning, expert systems, fuzzy inferencing and connectionist modeling based on artificial neural networks. The focus is on the use of soft computing approaches to deal effectively with real world complex systems for which their mathematical or physical models are either non-tractable or are difficult to obtain. The main thrust is on designing computationally intelligent systems with human like capabilities in terms of reasoning, learning and adaptation. Tools of computational intelligence could be used in a wide range of engineering applications involving real world problems such as in: planning problems, intelligent control, autonomous robotics, speech understanding, pattern analysis, network design, face recognition, communication systems to name a few.
Faculty: Nanyun (Violet) Peng
Alexa Fellow: I-Hung Hsu
Alexa Fellow Bio: I-Hung Hsu is a Ph.D. student in Computer Science at the University of Southern California (USC) working on speech and natural language research. His research interests include natural language processing, deep learning, artificial intelligence, and machine learning and its applications. Before joining USC, he received his B.S. degree in Electrical Engineering from the National Taiwan University (NTU) in Taipei, Taiwan. In June 2017, I-Hung was awarded first prize at the Intelligent Conversational Bot Final Competition, and in May 2016, he was awarded the Best Maker Award in the 2016 MakeNTU Hackathon. I-Hung recently coauthored a paper that was presented at the IEEE Automatic Speech Recognition & Understanding Workshop (ASRU), in December 2017.
Course description: USC is offering a new Human Language Technologies (HLT) certificate/specialization for the MS in Computer Science degree starting from the fall 2018 semester. This new program offers eligible MS students the opportunity to get deeper training in natural language and speech processing in the form of required courses in speech, dialog/NLP and machine learning. The specialization also requires two semesters of directed research focused on a speech/language research task.
Faculty: Sanjeev Khudanpur, Daniel Povey
Alexa Fellow: Vimal Manohar
Alexa Fellow Bio: Vimal Manohar is a PhD student in Electrical Engineering at Johns Hopkins University, where he works on general area of speech processing with Assistant Professor, Sanjeev Khudanpur, and Assistant Research Professor, Daniel Povey, in the Center for Language and Speech Processing in the. Manohar is focused on semi-supervised training and unsupervised adaptation of acoustic models for automatic speech recognition. His other research interests include acoustic models for robust speech recognition, speech activity detection and speaker diarization. He actively contributes to the open-source Kaldi speech recognition toolkit. Manohar earned a M.S. in Electrical Engineering from Hopkins, and a Bachelors in Electrical Engineering from Indian Institute of Technology, Madras.
Course description: The Machine Learning for Signal Processing course will focus on the use of machine learning theory and algorithms to model, classify, and retrieve information from different kinds of real-world, complex signals including audio, speech, text, image, and video. The course makes the link between signal processing and machine learning. Machine learning is a science which consists of developing powerful methods that are able to model and understand the different components of a given signal. Those methods rely on signal processing to process any kind of data. The course has a final project component that students work on over a full semester. These projects can be carried out using Alexa developer services.
Faculty: Mari Ostendorf
Alexa Fellow: Hao Fang
Alexa Fellow Bio: Hao Fang is a PhD candidate at the University of Washington (UW) advised by Prof. Mari Ostendorf. His research interests focus on social chatbots and natural language processing. He is a recipient of 2018 UW College of Engineering Student Research Award. In 2017, he led the UW Sounding Board team and won the inaugural Amazon Alexa Prize for building a socialbot that converses with human users on popular topics and recent news. He obtained his M.S. degree from the University of Alberta, and B.S. degree from Beijing University of Posts and Telecommunications.
Course description: Conversational Artificial Intelligence - The goal of this course is to introduce students to current methods and recent advances in conversational artificial intelligence (AI) and provide hands-on experience building a conversational AI system. The course covers basic components of a modular dialogue system as well as end-to-end systems, evaluation, and cloud-based implementation. The emphasis is on conversational (vs. task-oriented) systems. Students will work in multi-disciplinary teams on labs and a final project. Amazon Alexa and AWS will be used as the primary development platform.
Faculty: Hal Abelson
Alexa Fellow: Jessica Van Brummelen
Alexa Fellow Bio: Jessica is an MIT Computer Science graduate student aiming to empower young learners with technology to solve real-world problems. At the University of British Columbia, Jessica developed a cyclist collision warning system, researched autonomous vehicle technology in France, and investigated environmental engineering in Sweden. She also taught visual programming skills, developed micro-drone workshops, and tutored computer science. At MIT App Inventor, she is developing conversational artificial intelligence tools to allow anyone to create their own intelligent systems. She believes that given the right tools, knowledge and skills, anyone can create significant positive change in their community.
Course description: MIT's 6.S198 prepares students to develop deep learning projects. The first part of the course involves surveying basic techniques, including convolutional neural networks, recurrent neural networks, generative adversarial networks, and embedding. For each technique, fundamental concepts, open-ended demo applications, and programming assignments will be explored. The second part of the course involves designing and implementing original projects using these techniques. For example, a recurrent neural network may be used to develop a conversational AI application. Project teams will have mentors who are machine learning experts from industry. Related policy and social issues are also discussed.
Faculty: C V Jawahar
Alexa Fellow: Praveen Krishnan
Alexa Fellow Bio: Praveen Krishnan is a senior graduate student at IIIT Hyderabad. His interests are in the space that overlaps with computer vision and language processing. He has worked in the past on problems related to learning embedding/representation on visual data with semantic and linguistic interpretations. Presently, he also acts as a mentor in the course on Foundations of AI and ML.
Course description: Foundations of AI and ML - This course introduces the fundamental concepts in modern AI with equal emphasis on the practical aspects. Algorithms and solutions are introduced through examples and practical situations. Course follows a layered model of learning with practice-theory-practice. Beyond the lectures, this course also has laboratory sessions and mini projects that provides practical experience on working with real life problems and data.
Faculty: Junyi Jessy Li, Greg Durrett, Raymond J. Mooney
Alexa Fellow: Wei-Jen Ko
Alexa Fellow Bio: Wei-Jen Ko is a second year Computer Science PhD student at UT Austin advised by Prof. Junyi Jessy Li and Prof. Greg Durrett. He received his B.S. in Electrical Engineering from National Taiwan University in 2016. His research interests are in natural language processing and computer vision, with 5 publications at AAAI, ICASSP and other top venues. His prior research includes multi-task learning, deep canonical correlation analysis, and temporal ensembling, with applications to sentence specificity prediction, attention-based language modeling, and multi-label image classification.
Course description: CS378 - Natural Language Processing: This is an upper-division undergraduate natural language processing course covering properties of language (syntax, lexical and distributional semantics, compositional semantics, discourse, morphology) as well as fundamental machine learning concepts needed for NLP (log-linear models, sequence labeling, neural networks). Students will gain hands-on experience building and modifying real-world NLP systems including a text classifier, an information extraction system, a small-scale neural machine translation system, and a conversational dialogue system. Students will apply these models to a variety of domains and gain exposure to the process of data annotation.
Faculty: Dr. Andreas Vlachos
Alexa Fellow: James Thorne
Alexa Fellow Bio: James Thorne is a PhD student researching new ways in which Artificial Intelligence can be used to verify the truthfulness of information. After completing his Master of Engineering in Computer Science at the University of York, he started his PhD at Sheffield University with Andreas Vlachos. During this time, in conjunction with an internship at Amazon Research Cambridge, he developed the Fact Extraction and VERification (FEVER) dataset and shared task. James has transferred his work to the University of Cambridge where he will work in closer collaboration with Andreas and Amazon as part of the Alexa Fellowship.
Course description: Alexa devices will be used for seminar style teaching and graduate level projects in conversational AI along the following research directions: natural language generation with a focus on generation of passages consisting of multiple sentences; question generation with a focus on identifying questions humans ask when fact checking a claim; and analysis of conversations with a focus on assessment of markers of constructive dialogues. The students will be encouraged to use Alexa for both demonstrating the methods developed as well as interacting with their users.
Faculty: Dr. Nikolaos Aletras, Prof. Eleni Vasilaki
Alexa Fellow: TBA
Alexa Fellow Bio: TBA
Course description: The series will introduce the Alexa Skills Kit to Undergraduate & Masters and PhD students undertaking modules in language and speech technology. The students will develop Alexa skills to utilize the Amazon’s voice service. Projects will include the development of Alexa skills and students will require to use technologies such as AWS Lambda. We will organize a one-day hackathon where undergraduate and postgraduate students will come up with new Alexa skills that will implement during the day.
Only PhD students and post-doctoral fellows at universities invited by Amazon to participate in the Alexa Graduate Fellowship are eligible. The 2018-2019 class has been selected and we are not accepting applications at this time. Sign up below for our interest mailer to stay informed about the program and any future call for applications.
The Alexa Graduate Fellowship includes:
The Alexa Graduate Fellowship is intended to fully cover tuition and stipend for Graduate Fellows, though this amount varies across participating universities. AWS credits are available to students based on the variable research needs of the specific project.
No, all intellectual property resulting from the work of an Alexa Graduate Fellow will belong to the Graduate Fellow in compliance with any university policies. Amazon takes no ownership in any intellectual property, and encourages all Graduate Fellows to publish their work, contribute to the open source community, or pursue protection of their intellectual property as they see fit.
Amazon works directly with faculty at the invited universities whose research groups are focused on fields relating to the advancement of spoken language systems, including but not limited to: wake-word detection, audio signal processing, automatic speech recognition, speaker diarization, voice recognition, natural language understanding, entity resolution, contextual reasoning, dialogue management, question answering, and text-to-speech natural language generation. Alexa Fellows are selected by these faculty advisors to pursue research questions of their choosing within a field related to the focus of the research group.
Yes, Alexa Fellows are welcome to apply for an internship at Amazon, though this is neither required nor guaranteed.