RMSDNet: A Lightweight Object Detection Network for Rail Surface Defect
Bibliographic record
Abstract
The surface condition of rails is important for ensuring the safe and stable operation of railway vehicles, so real-time defect detection of rail surfaces is essential. However, manual inspection and mainstream non-destructive surface detection methods are not only difficult to meet the accuracy requirements but are also inefficient. To solve this problem, we propose a new rail surface defect detection method, namely, reversible multi-scale detection networks (RMSDNet) based on the improved YOLOv8-n, which can detect rail surface defects more accurately and quickly with fewer parameters and greater efficiency. First, the backbone is reconstructed using the concept of reversible column networks (RevCol) to complete feature extraction more efficiently. Secondly, the multi-section block with attention and pooling (MSAP) module is designed to enhance attention to defects and reduce noise interference during feature fusion. In addition, ghost convolution with shuffle (GSConv) is introduced to reduce the computational complexity in the process of down-sampling and further optimize the information interaction. Finally, a semi-decoupled head (SD-Head) is designed to reduce the information redundancy while ensuring detection accuracy. Experiments on the rail surface defect dataset show that our model achieves the highest mAP@0.5 of 78.0% with the fewest parameters and lowest FLOPs compared to other mainstream object detection models.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".