TheMachine Press

A daily newspaper for the age of artificial intelligence.

Morning editionPermanent story

research

Three Labeled Days Estimated Salmon Freshness

A few-shot hyperspectral model reached a 1.58-day mean error on unseen fillets in a 16-day dataset.

Published Updated Story ID: mp-2026-08-13-017
Read the complete editionStory JSON

Summary

A few-shot hyperspectral model reached a 1.58-day mean error on unseen fillets in a 16-day dataset.

With three labeled days per fillet, the ordinal model reported 72.3 percent accuracy within two days. The test was a controlled hyperspectral dataset and estimates storage age, not food safety or edibility.

Why it matters

A few-shot hyperspectral model reached a 1.58-day mean error on unseen fillets in a 16-day dataset.

Limits and context

  • The test was a controlled hyperspectral dataset and estimates storage age, not food safety or edibility.

Key claims

  1. A few-shot hyperspectral model reached a 1.58-day mean error on unseen fillets in a 16-day dataset.

    Qualification: The test was a controlled hyperspectral dataset and estimates storage age, not food safety or edibility.

    Evidence: source-2026-08-13-019

Sources

  1. arXiv preprint 2608.12230arXiv · primary research

Corrections

No corrections have been recorded for this story.