Aero-engine combustion flame segmentation via Vision Transformer
Bibliographic record
Abstract
Understanding and monitoring flame behavior in aero-engine combustors is critical for ensuring safe and efficient operation. The turbulent and high-pressure environment induces complex flame topologies and thermoacoustic instabilities that challenge conventional diagnostics. Traditional intensity-based thresholding methods are often inadequate for capturing coherent flame structures. In this study, we present the deep learning framework FlameSeg designed to improve segmentation accuracy and enable high-fidelity kinematic analysis of turbulent flames. FlameSeg integrates two key components: (1) hierarchical feature extraction with attention mechanisms to capture both fine boundary details and global contextual information, ensuring that subtle flame structures are preserved; and (2) a lightweight decoder with multi-scale feature fusion, which effectively integrates information across multiple resolutions, enabling precise delineation of flame contours and robust representation of overall flame structures. For validation, we established FlameDataset, a dedicated high-speed imaging collection from a representative aero-engine combustor. On this dataset, FlameSeg achieves a state-of-the-art mean Intersection over Union of 92.41%. The derived flame centroid trajectories attain an average mean absolute error of 8.07 pixels and a phase-averaged normalized error Enorm of 3.06% relative to the flame's characteristic diameter. This precision allows for the resolution of subtle kinematic variations often associated with the onset of combustion instabilities. These results demonstrate that FlameSeg constitutes a high-fidelity diagnostic framework for resolving turbulent flame dynamics, offering a robust pathway toward an improved understanding and monitoring of combustion instabilities.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".