Can AI See You? Your Introduction to Computer Vision and GenAI
summary
Starts from pixels and RGB then shows concrete pipelines: convolutional nets for classification, bounding boxes and pixelwise segmentation, real‑time face landmarks and pose tracking that run on a laptop CPU. Live demos use YOLO and MediaPipe, and a latent‑diffusion generator (Stable Diffusion) that inpaints and reimagines selfies. A working AI photo booth and a deep‑fake face‑swap showcase are built and demonstrated, models are task‑specific or VLM‑conditioned, and the demos reliably run locally.
AI-written summary · about
chapters (14)
- 0:00AI and computer vision intro
- 4:20How computers see images
- 6:50Convolutional neural networks
- 12:29Reconstructing images and deepfakes
- 14:38Object detection and segmentation
- 17:51Facial landmark detection
- 20:24Vision language models and GenAI
- 24:18Text to image models
- 34:18AI photo booth applications
- 36:04Summary: AI modifies images
- 37:29Q&A: Autonomous driving
- 40:20Vision transformers vs CNNs
- 41:47AI and SEO for visuals
- 42:31Quality inspection models
