CSU-AWS Grant 2018: AI-Powered Accessible Streaming

Accessible Instructional Video Content Streaming is a collaborative project of two groups in the Technology Services organization (the Hub for Innovation, Research, and Development; and Academic Technology), and faculty research partners. Together we  are discovering and adopting emerging innovations for advancements in education with Cloud technology and Artificial Intelligence.

Captioning isn’t just a good idea, it’s required! Also, the benefits of captioning go beyond accessibility. Katie Linder of Oregon State University says “about 75% of student qualitative comments that talked specifically about learning benefits“ (“The State of Closed Captioning in Higher Education”). Currently, Academic Technology captions a limited number of instructional video presentations without charge to faculty through funding by Academic Affairs. We believe that Automatic Speech Recognition (ASR) technology is  potentially a cost effective approach of producing transcriptions with acceptable accuracy.

This project is to develop and implement an on-demand digital repository with Amazon CloudFront to deliver instructional video content in HTTP Live Streaming (HLS) Mobile-ready format; and, with Amazon Transcribe – an automatic speech recognition (ASR) service, to add speech to text capability for Accessibility – Closed Captioning. Other AWS Cloud services will also be adopted, i.e. S3 for video storage and Amazon Glacier for archiving. There will also be a front end UI for faculty to upload and manage videos and a dashboard for viewing statistics.

This project aligns closely with the Cloud Infrastructure Readiness effort at Fresno State.

The successful development of the project will support several areas of Fresno State Strategic Plan 2016-2020:

  • Embrace technology to support innovation, teaching and learning. (Criteria: Transformational)
  • Enhance teaching and learning through innovative programs toward the CSU’s central strategic objective – Graduation Initiative 2025 (Criteria: Enterprise Impact)
  • Through hiring student intern developers and the course research work designed by the faculty researchers, support HIPs of Undergraduate Researches and Internship (Criteria: Research and Education)

About CSU-AWS Credits Grant | Project Updates:  7/2/2018; 10/1/2018; 1/2/2019; 4/1/2019; 7/1/2019

Dear Colleagues,

Congratulations! Your proposal has been selected for funding for AWS credits.

If you have any questions about the overall program, please reach out to me. Paul will coordinate the accounting for the credits so questions regarding that aspect can be addressed to him. He can also help connect you to resources at Cal Poly or AWS that can assist you with planning and analysis regarding AWS services and campus readiness.

Thank you so much for submitting a winning proposal, and I look forward to seeing what you can accomplish as a model for the CSU and beyond.

Michael

Summary Project Plan, including milestones and dates. Note: Projects approved in this round must be completed fully no later than February 28, 2019.

This project is proposed in three phases:

Phase one of this project will emphasize on technology research and development. A basic AWS infrastructure consisting of the test, development, and production ROOT accounts with Single Sign On access (Shibboleth) will be established before the end of Q1 2018 including accesses to all AWS services (listed in the project description) and IAM accounts for faculty researcher, student developers and test users are provisioned; and, Initial test content (i.e., streaming videos) and data (i.e., meta data) should be loaded (in S3 or RDB). Project design and development will kick in Q2 and Q3 to complete the prototypes (concept proving) for pilots.

Phase two of this project will involve students/classes for pilot testing during Fall 2018 semester.

Phase three (Jan/Feb 2019) to complete the project evaluation, assessment review, documentation, and grant report.

Assessment Plan. Explain how you will evaluate your project and its outcomes

  • User Testing by piloting with student participants in live classes by Pre- and post-semester questionnaires.
  • Cost Analysis: Between the current Automatic Sync and AWS Transcribe.
  • Quality Assessment: Faculty Resource center will provide student assistants for manually checking on the accuracy of generated transcriptions.
  • Technical Feasibility study.

Sustainability Plan. Explain how any ongoing or future work that results from this project will be funded in future years. 

The successful outcome of this project is to improve a current service with cost saving through technology innovation. The project outcome and study will be presented to the Technology Services and Academic Affairs leaderships for the consideration of new system adoption to replace the Automatic Sync service.

Dissemination Plan. Explain how you will share the results of your work with other CSU’s and beyond. 

  • Share with CSU peer campuses for future collaboration and establish partnerships
  • Present in the national conference, i.e., Educause, AWS re:Invent
  • Apply grants for continue development or large scale research studies
  • Publishing the design and code development with the Open-source community.

For additional project information, please contact Max J. Tsai.

ITAC Sponsor:
Orlando Leon, CIO

Lead (PI)
Max J. Tsai, Innovation Architect

Project Sponsors:
Dr. Brent Auernheimer, Director of Academic Technology (Exec. Sponsor)
Brad Barker, Director of Infrastructure and Systems

Key Personnel:
Jay Fowler, Systems – Technical Lead
Mike Bach, Academic Technology – LMS and Resource Center Lead

Research Collaborators:
Dr. Alex Liu, Innovation Faculty Fellow