Ethical Practices for Collecting Ground-Truth Food Datasets: A Systematic Review
Bibliographic record
Abstract
Food image dataset collection for artificial intelligence (AI) studies has ethical concerns that are underestimated. Thousands and millions of images are collected from different sources on the Web, with ethical challenges relating to copyright. While it is difficult to collect large amounts of food images based on permissions, a small number of datasets can affect the performance of the machine learning model (MLM). It becomes imperative to consider a balance between ethics, data collection, and machine learning model performance. We present state-of-the-art ethical considerations for collecting food image datasets. A total of 102 papers were reviewed, and we found that only a few papers (4) reported on the ethical practices they adopted for collecting food image datasets. These ethical practices include obtaining permissions from data sources such as websites and social media sites and obtaining ethical approval for collecting food datasets from participants in food logging studies. For future work, we present opportunities, challenges, and recommendations for considering dataset collection in the food domain. Though there are challenges around collecting datasets that are sufficient for training MLMs, we provide recommendations to balance the trade-off between gathering large datasets ethically and improving the accuracy of MLMs.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.038 | 0.141 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.004 | 0.004 |
| Bibliometrics | 0.009 | 0.007 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.003 | 0.005 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.003 | 0.002 |
| Insufficient payload (model declined to judge) | 0.005 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".