Robustness Assessment of AI-Based 2D Object Detection Systems: A Method and Lessons Learned from Two Industrial Cases
Bibliographic record
Abstract
The reliability of AI-based object detection models has gained interest with their increasing use in safety-critical systems and the development of new regulations on artificial intelligence. To meet the need for robustness evaluation, several authors have proposed methods for testing these models. However, applying these methods in industrial settings can be difficult, and several challenges have been identified in practice in the design and execution of tests. There is, therefore, a need for clear guidelines for practitioners. In this paper, we propose a method and guidelines for assessing the robustness of AI-based 2D object detection systems, based on the Goal Question Metric approach. The method defines the overall robustness testing process and a set of recommended metrics to be used at each stage of the process. We developed and evaluated the method through action research cycles, based on two industrial cases and feedback from practitioners. Thus, the resulting method addresses issues encountered in practice. A qualitative evaluation of the method by practitioners was also conducted to provide insights that can guide future research on the subject.
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 | 0.018 | 0.040 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.001 | 0.004 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.003 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
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".