Multimodal AI
Vision model
A model designed to understand or generate information from images or video.
Example
A vision model identifies damaged parts in photographs from a production line.
Why it matters
It helps people examine visual material at a scale that would be slow to review manually.
You'll hear it
“It recognized the object but missed the small warning label.”
Can you control it?
Sometimes
Sometimes. Some products or workflows let you influence a vision model, while others keep it fixed or manage it automatically. Check the settings and documentation for the tool you are using.
Common questions
- Does a vision model see exactly as people do?
- No. Its learned patterns can fail in ways that seem surprising to a person.
- Can lighting change the result?
- Yes. Shadows, glare or poor image quality can affect what it recognizes.
- Does visual recognition establish what happened before the picture?
- No. The image may not provide enough evidence about the wider event.