Digital Media Accessibility, Captioning and Transcription Services

The Vendor is required to provide digital media accessibility, captioning and transcription services for include:
- Platform and integration
1. Vendor platform features
•    A secure, role-based web portal for administrative functions (manual job submission, tracking, and billing). 
•    Tracking should provide granular, auditable reports on usage by colleges, and departments to facilitate accurate cost recovery, bill-back, and budget projection. 
•    The portal should also allow for centralized management of user access and permissions across the multi-institutional contract.
- Enhanced accessibility features
1. Emerging platform and XR capability
•    Capability to produce caption files and delivery mechanisms compatible with virtual reality (VR), augmented reality (AR), and extended reality (XR) headsets and immersive instructional platforms (e.g., unity, unreal engine, Meta quest).
2. Audio description player
•    Provision of a high performance, resource-efficient, accessible, and embeddable media player specifically designed to fully support audio description (ad) and extended audio description (EAD) playback for use on institutional web pages and course management systems to mitigate the current lack of native ad support in many common third-party video platforms.
3. AI audio description (AI-ad) capability and AI-ad quality control
•    Provide an AI-assisted audio description service that utilizes machine learning and computer vision to generate the ad script and synthetic voice narration.
•    For all AI-ad content, should provide a human-in-the-loop (HIL) review process to ensure the final product meets the following quality standards: 
o    Accuracy - the description must be factually correct and true to the visual information presented on screen (objects, text, locations, and actions).
o    Clarity - descriptions must be concise, unambiguous, and use simple, active language to maximize understanding within limited time gaps. 
o    Objectivity - descriptions must be strictly neutral and factual, detailing only what is seen, not what is inferred, interpreted, or subjectively felt, and
o    Correct timing - the description must be precisely timed to fit into the natural audio pauses and must not overlap any spoken dialogue or essential sound effects.
4. Synthetic voice
•    Must offer a selection of high-quality synthetic voices (text-to-speech) for ad narration that are customizable, natural-sounding, and distinguishable from the main program audio.
- Technology and language
1. AI and ASR utilization
•    Use of advanced automated speech recognition (ASR) technology, combined with a human review and editing process (HIL), to optimize turnaround time while maintaining   99% accuracy. 
2. Multilingual capabilities
•    Ability to provide subtitling and translation services for non-English content, with delivery in all required foreign-language formats and using UTF (Unicode transformation format)-8 encoding for full character support.