Apps That See 👀 Practical AI Vision for Developers
summary
Demos show open source apps that process images and video using REA APIs and local models like Gemini 3 4B on a MacBook Pro; repos are shared via QR. Specific builds include a media library that swaps backends (and normalizes differing bounding‑box formats), a video→blog pipeline that extracts keyframes with ffmpeg and generates structured posts, a searchable video index that produces Q&A and quizzes, and a no-code n8n clipper that auto-creates captioned social clips; latency, token cost, and local vs cloud privacy tradeoffs are measured.
AI-written summary · about
chapters (18)
- 0:00AI vision apps intro
- 1:04What apps that see can do
- 2:30Why AI vision is accessible
- 4:57Local vs cloud vision AI
- 7:07Practical AI vision uses
- 9:17Demo overview and projects
- 10:43Generating image prompts
- 13:36Comparing vision AI models
- 15:45Media library analyzer
- 17:32Describing images locally
- 19:50Object detection in videos
- 21:44Videos to blog posts
- 25:13Searching video content
- 26:57Q&A from video tutorials
- 28:46AI visual context
- 30:53No-code video automation
- 31:55Terminal AI vision demo
- 33:20Q&A on costs
related talks
44mThe age of applied AI
01h 03mBuild & deploy AI-powered apps
18mBuilding AI-powered video generation tools for web and mobile platforms
42mAI in React JS
25mPersonalizing Your Images with AI-Powered Features
11mAI-powered JavaScript applications
33mSee More, Be More
27mDevelop Computer Vision Apps With Go