Concept
Mixup
Mixup是一种机器学习领域的数据增强技术，由Zhang等人于2018年提出。其核心思想是通过对训练样本及其对应标签进行线性插值，生成新的虚拟训练样本，从而扩充训练数据集。Mixup能够有效提升模型的泛化能力、增强对对抗样本的鲁棒性，并缓解过拟合问题，广泛应用于图像分类、自然语言处理等多种深度学习任务中。
Timeline (last 90 days)
Sep 11
Training sets can be expanded by leveraging data augmentation techniques (such as back-translation and Mixup) or generative models (such as diffusion models and GANs).
Unverified50%
Aug 25
Mixup和CutMix等混合增强方法在暗条分割中需谨慎使用，因为丝状形态容易在混合中丢失语义一致性
Unverified50%
All Facts (3)
Unverified
The Mixup technique was proposed by Zhang et al. in 2018, with its core operation being linear interpolation of training samples.
70%UnverifiedTraining sets can be expanded by leveraging data augmentation techniques (such as back-translation and Mixup) or generative models (such as diffusion models and GANs).
50%UnverifiedMixup和CutMix等混合增强方法在暗条分割中需谨慎使用，因为丝状形态容易在混合中丢失语义一致性
50%