AI/ML demo derby

on YouTube ↗17m

" summary

Three concise demos show working pipelines and numbers not promises: a pipeline uses Cloud Video Intelligence API label timestamps plus the Transcoder API overlay injection to produce a “phobia‑safe” video that actually hides detected swans; an Edge TPU/Coral + Vertex AI model exports in TensorFlow Lite, streams simple occupancy counters into BigQuery and Looker; a PDF→audio pipeline runs Vision OCR, classifies paragraphs with AutoML Tables (95% body / 97% garbage accuracy) then calls Text‑to‑Speech to emit MP3s in ~10–15 minutes.

AI-written summary · about

" chapters (8)

1. 0:00Start
2. 0:28Backstory
3. 1:10Cloud Video Intelligence Api
4. 1:34The Transcoder Api
5. 12:40Vision Api
6. 13:10Use Vision Api
7. 15:32Evaluation Report
8. 16:00Text to Speech Api

" related talks

• 01h 03mIt's All About AIJonas Mayer, Martin Förtsch, Thomas Endres · YOW! Melbourne 2022
• 13mPremiereConf42 Machine Learning 2021
• 47mBuilding smarter apps with Machine Learning, from magic to realityLaurent Picard · GDG DevFest Poland 2018 in Wrocław
• 14mMachine Learning for developers in a hurryDale Markowitz · Google I/O 2022
• 57mBuilding smarter solutions with Machine Learning, from magic to realityLaurent Picard · code.talks commerce 2019
• 50mML and AI at the edge with Google Coral, EdgeTPU and Cloud PlatformMarkku Lepisto, Grant Likely, Tamas Daranyi · Linaro Tech Days 2020
• 40mVertex AI: Pipelines for your MLOps workflowsMarton Kodok · Voxxed Days Luxembourg 2022
• 22mBoring but Effective: WebAI That ShipsNico Martin · dotJS 2026