Short Answer

Why can examples from one photo source help train on another?

Question: In one to three sentences, explain why product photos taken in a studio can still help train a model that will be used on customer-uploaded phone photos.

Sample answer: Because studio photos and phone photos still share many visual patterns, such as object shapes, colors, and textures, the model can reuse features learned from the studio images when processing phone photos. That shared structure makes the extra studio data useful even though the two sources are not identical.

Key points:

  • Shared visual patterns across the two photo sources
  • Feature knowledge learned from one source can transfer to the other

Rubric: Full credit for stating both the shared patterns and the transfer of learned features between the two sources; partial credit for only saying that more examples help without explaining why the transfer works.

0

1

Updated 2026-08-12

Contributors are:

Who are from:

Tags

Machine Learning

Deep Learning

Supervised Learning

Dive into Deep Learning @ D2L

Data Science

Machine Learning Strategy

Machine Learning Yearning @ DeepLearning.AI