Improved Metric-Learning-Based Recognition Method for Rail Surface State With Small-Sample Data

The accurate identification of rail surface states, especially the third media state, is crucial for enhancing traction and braking capabilities of heavy haul trains, ensuring their safe operation, and maintaining heavy haul railways. Few-shot learning is commonly used to recognize rail surface states, effectively addressing the overfitting issue caused by limited sample data. However, in practical rail surface state data situations, few-shot learning faces challenges such as insufficient extraction of crucial feature information and a tendency to lose distinguishing degree information. To address these challenges, this paper proposes a rail surface state recognition model based on improved metric learning. The proposed method incorporates a pyramid-splitting attention mechanism in the feature extraction network. This allows for the extraction of multi-scale spatial information from the feature map, facilitating cross-dimensional channel attention and interaction between spatial attention features. This addresses the issue of inadequate key feature information extraction caused by a limited number of orbital surface state samples. Additionally, a deep local description concatenator splices the local features of the query set and various support set feature maps in pairs, replacing the global feature splicing in traditional metric learning. This enables the filtering of interference information, such as background, while retaining feature information with significant differentiation to a larger extent. The proposed method was evaluated using a small-sample rail surface state dataset that we constructed. According to the experimental results, the proposed method outperforms existing methods in terms of recognition accuracy, precision, and recall.