Vision and Multimodal (VLMs)
Vision and Multimodal (VLMs)
Overview
Computer vision models understand images or video. VLMs process both text and images.
Tasks
- Classification, detection, segmentation, OCR
Example
- Use an OCR model to extract text from receipts and parse totals
Checklist
- Consider input sizes and latency
- Test on your real data (lighting, angles)
Resources
- OpenCV
- Ultralytics YOLO
- OpenAI/Anthropic VLM model docs