- Updated: August 23, 2026
- 2 min read
Few-Shot Ordinal Learning for Day‑Wise Freshness Estimation with Hyperspectral Fish Images
Few-Shot Ordinal Learning for Day‑Wise Freshness Estimation with Hyperspectral Fish Images
Authors: Kazi Nabiul Alam, Pooneh Bagheri Zadeh, Akbar Sheikh‑Akbari
Non‑destructive food quality assessment has been revolutionised by hyperspectral imaging (HSI), which captures detailed spectral signatures linked to biochemical changes during storage. While HSI excels at detecting freshness, estimating day‑wise freshness remains challenging due to strong inter‑fillet variability and the scarcity of labelled data for each product.
Introducing Few‑Shot Ordinal Learning
We present the first few‑shot learning framework tailored for HSI‑based food quality estimation. Each salmon fillet is treated as an independent episodic task, and a CORAL‑style ordinal prediction head models the ranked nature of freshness progression through cumulative threshold modelling. Biologically‑grounded monotonicity and embedding smoothness constraints guide predictions toward plausible temporal trajectories.
Key Results
- Mean Absolute Error (MAE): 1.58 days
- 2‑day accuracy: 72.3 %
- Only three labelled days per fillet required
These results substantially outperform scalar regression and label‑distribution baselines under an unseen‑fillet protocol on a 16‑day salmon HSI dataset.
Why It Matters
The proposed method dramatically reduces the annotation burden while delivering high‑precision freshness estimates, enabling scalable, real‑time monitoring in commercial supply chains.
Read the Full Paper
For a detailed description, methodology, and experimental setup, visit the internal paper page: https://ubos.tech/papers/2608.12230.
Explore related research and tools on our site: https://ubos.tech/research and https://ubos.tech/tools.
Stay tuned for upcoming releases and applications of few‑shot ordinal learning in other food‑quality domains.
Andrii Bidochko
CTO UBOS
Andrii Bidochko is an AI entrepreneur and researcher focused on AI agents, reinforcement learning, and autonomous systems. He writes about the technologies shaping the future of machine intelligence, from frontier models and agent architectures to real-world AI applications.