Abstract
Real-time semantic segmentation presents the dual challenge of designing efficient architectures that capture large receptive fields for semantic understanding while also refining detailed contours. Vision transformers model long-range dependencies effectively but incur high computational cost. To address these challenges, we introduce the Large Kernel Attention (LKA) mechanism. Our proposed Bilateral Efficient Visual Attention Network (BEVANet) expands the receptive field to capture contextual information and extracts visual and structural features using Sparse Decomposed Large Separable Kernel Attentions (SDLSKA). The Comprehensive Kernel Selection (CKS) mechanism dynamically adapts the receptive field to further enhance performance. Furthermore, the Deep Large Kernel Pyramid Pooling Module (DLKPPM) enriches contextual features by synergistically combining dilated convolutions and large kernel attention. The bilateral architecture facilitates frequent branch communication, and the Boundary Guided Adaptive Fusion (BGAF) module enhances boundary delineation by integrating spatial and semantic features under boundary guidance. BEVANet achieves real-time segmentation at 33 FPS, yielding 79.3% mIoU without pretraining and 81.0% mIoU on Cityscapes after ImageNet pretraining, demonstrating state-of-the-art performance. The code and model is available at this https URL.
Abstract (translated)
实时语义分割面临着既要设计高效架构来捕获大感受野以进行语义理解，又要细化详细轮廓的双重挑战。视觉变换器能够有效地建模长距离依赖关系，但会带来较高的计算成本。为了解决这些问题，我们引入了大型内核注意力（LKA）机制。我们的提出的双边高效视觉注意网络（BEVANet）扩展感受野以捕获上下文信息，并使用稀疏分解的大可分离内核注意（SDLSKA）提取视觉和结构特征。综合核选择（CKS）机制动态调整感受野，进一步提升性能。此外，深度大型内核金字塔池化模块（DLKPPM）通过结合膨胀卷积和大型内核注意力来丰富上下文特征。双边架构促进了频繁的分支通信，并且边界引导自适应融合（BGAF）模块在边界指引下整合空间和语义特征以增强轮廓定义。 BEVANet实现了每秒33帧的速度，在不进行预训练的情况下，mIoU达到了79.3%，而在Cityscapes数据集上经过ImageNet预训练后，mIoU达到81.0%。这显示了它在实时语义分割任务中的卓越性能。代码和模型可以在提供的链接中获取。
URL
https://arxiv.org/abs/2508.07300
