A system must find and locate every car in a street photo. Which task is this?
Why B is the answer
Object detection
Object detection finds objects and where they are. Classification gives a label for the whole image.
In detail
- Object detection finds and locates each item in an image.
- Classification names the image, and detection locates things in it.
Use case
An illustrative scenario, not a real incident.
A city counts cars in street photos. Object detection finds each car and draws a box around it, which classification of the whole photo could not do.
See it in different ways
flowchart TD A["Image"] --> B["Classification: what is this?"] A --> C["Object detection: what and where?"] A --> D["OCR: what does it say?"]
mindmap
root((AI workloads))
Text
Sentiment
Key phrases
Entities
Speech
Recognition
Synthesis
Vision
Classification
Object detection
OCR
Documents
Information extraction
Generative
Text
Imagesflowchart TD
Q{"What is the output?"}
Q -->|"One label for the photo"| A0["Image classification"]
Q -->|"Each object and its position"| A1["Object detection"]Key point: Finding and locating is object detection.
Why the other options are wrong
- A. Image classification of the whole photo
- Classification labels the whole photo, without locations.
- C. Speech recognition
- Speech recognition works on audio.
- D. Sentiment analysis of the image caption
- Sentiment analysis works on text.