Abstract: Data augmentation is a series of techniques that generate high-quality artificial data by manipulating existing data samples. By leveraging data augmentation techniques, AI models can ...

Abstract: Data augmentation is crucial for addressing insufficient training data, especially for augmenting positive samples. However, existing methods mostly rely on neural network-based feedback for ...

Time series classification is widely used in many fields, but it often suffers from a lack of labeled data. To address this, researchers commonly apply data augmentation techniques that generate ...