MD4K
收藏资源简介:
MD4K数据集是由北京航空航天大学虚拟现实技术与系统国家重点实验室创建的小规模训练集,包含4172张图像,旨在减少对大规模训练数据的依赖。该数据集通过整合MSRA10K和DUTS-TR数据集的优点,实现了语义类别的平衡。MD4K数据集不仅提高了现有模型的性能,还为研究者提供了一种构建训练集的新方法。该数据集的应用领域主要集中在显著对象检测,旨在解决模型训练中对大规模数据集的依赖问题。
The MD4K dataset is a small-scale training set developed by the State Key Laboratory of Virtual Reality Technology and Systems, Beihang University. It contains 4172 images and aims to reduce the reliance on large-scale training data. This dataset achieves balanced semantic categories by integrating the strengths of the MSRA10K and DUTS-TR datasets. The MD4K dataset not only enhances the performance of existing models but also provides researchers with a novel approach to constructing training sets. Its application scenarios primarily focus on salient object detection, targeting the problem of excessive dependence on large-scale datasets during model training.




