Maximize lifetime value of media content
summary
Demonstrates a working pipeline that extracts millisecond metadata with Amazon Transcribe plus Rekognition Comprehend and Translate to enable search, compliance, ads and localization. Captioning platform produces frame-accurate captions, reaches 1,000,000+ minutes/month, and returns ASR at ~97–98% accuracy then human edits to 100%. Customers cut promo production from two days to two hours, one cuts subtitle time to an eighth, and live streams get near real time translation into 25 languages delivering up to 12x ROI.
AI-written summary · about
chapters (14)
- 0:00Introduction and speakers
- 0:45Current media trends and needs
- 4:02Why AI and ML in media workflows
- 7:21AWS Media Intelligence Solutions
- 10:27Four main use cases overview
- 10:47Search and discovery use case
- 13:19Compliance and brand safety
- 14:22Content monetization strategies
- 15:30Subtitling and localization
- 15:54CaptionHub platform overview
- 21:25CaptionHub benefits and integrations
- 23:55Live subtitling demonstration
- 31:48Next steps and resources
- 32:53Q&A and feedback
related talks
29mUsing Amazon Transcribe to make content searchable and accessible
31mShare information by removing language barriers with Amazon Translate
45mTranscend Language Barriers with Custom Machine Learning APIs: um...
30mWhich episode? Using AI to streamline media content operations
28mEnhance speech recognition accuracy using custom language models
27mSummarizing & analyzing videos
20mMeta's AI Translations Inference
24mApplications of AI in news and media