We introduce SANDesc, a Streamlined Attention-Based Network for Descriptor extraction that aims to improve on existing architectures for keypoint description. Our descriptor network learns to compute descriptors that improve matching without modifying the underlying keypoint detector. We employ a revised U-Net-like architecture enhanced with Convolutional Block Attention Modules and residual paths, enabling effective local representation while maintaining computational efficiency. We refer to the building blocks of our model as Residual U-Net Blocks with Attention. The model is trained using a modified triplet loss in combination with a curriculum learning-inspired hard negative mining strategy, which improves training stability. Extensive experiments on HPatches, MegaDepth-1500, and the Image Matching Challenge 2021 show that training SANDesc on top of existing keypoint detectors leads to improved results on multiple matching tasks compared to the original keypoint descriptors. At the same time, SANDesc has a model complexity of just 2.4 million parameters. As a further contribution, we introduce a new urban dataset featuring 4K images and pre-calibrated intrinsics, designed to evaluate feature extractors. On this benchmark, SANDesc achieves substantial performance gains over the existing descriptors while operating with limited computational resources.

翻译：我们提出了SANDesc，一种基于注意力机制的简化描述符提取网络，旨在改进现有的关键点描述架构。我们的描述符网络通过学习计算描述符来提升匹配性能，而无需修改底层关键点检测器。我们采用改进的类U-Net架构，通过卷积块注意力模块和残差路径进行增强，在保持计算效率的同时实现有效的局部表征。我们将模型的基本构建单元称为带注意力的残差U-Net块。该模型使用改进的三元组损失结合课程学习启发的困难负样本挖掘策略进行训练，从而提升了训练稳定性。在HPatches、MegaDepth-1500和Image Matching Challenge 2021上的大量实验表明，在现有关键点检测器基础上训练SANDesc，相比原始关键点描述符能在多项匹配任务中获得更好的结果。同时，SANDesc的模型复杂度仅为240万个参数。作为进一步贡献，我们提出了一个包含4K图像和预标定内参的新型城市数据集，专为评估特征提取器而设计。在此基准测试中，SANDesc在有限计算资源下运行，仍能较现有描述符实现显著的性能提升。