A Novel Segmentation Network for Multi-Target Segmentation in Track Scenarios

This study presents a novel segmentation network specifically designed for multi-target segmentation within track environments. We train and test our network on various datasets, including commonly used segmentation networks from recent years. Through rigorous experimentation, we demonstrate that our proposed segmentation method outperforms existing methods in accurately segmenting multiple targets within track scenarios.

The key contributions of this paper are as follows:

• Novel Segmentation Network: We propose a novel segmentation network specifically designed for multi-target segmentation in track scenarios.
• Improved Accuracy: Our network achieves superior accuracy compared to existing segmentation methods, demonstrating its effectiveness in accurately segmenting multiple targets within track environments.
• Effective Training and Testing: We extensively train and test our network on diverse datasets, including commonly used segmentation networks from recent years, to validate its performance.