Reality and beyond
summary
Blippar demonstrates a real‑time visual AI engine that classifies objects with a 33,000‑word vocabulary and claims 70–80% accuracy on general scenes, identifies car make and year in ~100 ms, and recognizes public figures from magazine photos; demos run live over 3G/4G and sometimes lag but mostly work, show gyro‑based interior views and AR overlays like Snoop Dogg t‑shirts, and emphasize semantic classification and spatial geometry as separate strands being combined.
AI-written summary · about
chapters (7)
- 0:00Introduction to AR and VR
- 1:02Differences between AR and VR
- 3:10AI and AR evolution
- 7:53Semantic vs geometry in AR
- 9:15Live AI visual demo
- 14:22Augmenting human figures
- 17:13Closing remarks
related talks
38mAI + AR Superpowers!
44mThe age of applied AI
46mBuilding accessible experiences with AI
01h 09mAI is eating the world (and Google)
44mUnderstanding the role of Artificial Intelligence in the present and in the future
43mThe Future of AI
21mAI Isn't Just Chat
36mApps That See 👀 Practical AI Vision for Developers
