New DGM-Net model offers efficient semantic segmentation with geometric guidance

Researchers have developed DGM-Net, an efficient architecture for semantic segmentation that bypasses the need for large models and high computational budgets. The network utilizes a novel Directional Geometric Mamba (G-Mamba) operator, which offers linear complexity for context modeling. By incorporating geometric guidance through centripetal flow fields and topological skeletons, DGM-Net enhances boundary preservation and achieves strong performance on benchmarks like Cityscapes and ADE20K, even under constrained hardware conditions. AI

影响 Introduces a resource-efficient architecture for semantic segmentation, potentially enabling wider deployment on edge devices.

排序理由 This is a research paper detailing a new architecture and operator for semantic segmentation.

在 arXiv cs.CV 阅读 →

• ADE20K
• arXiv
• Cityscapes
• DGM-Net
• G-Mamba

• 论文
• 其他

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →