Feature_Drop Data Augmentation in Audio and Speech (Feature-Drop Aspect) Paper: SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition (INTERSPEECH 2019) Time Masking Time Masking 최대 Time mask 길이를 t: Time mask 길이 t0 : Time mask 시작지점 Sampling 방법은 Time Masking 알고리즘과 동일하... DataAugmentationML/DLFeature_DropDataAugmentation
Data Augmentation in Audio and Speech (Feature-Drop Aspect) Paper: SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition (INTERSPEECH 2019) Time Masking Time Masking 최대 Time mask 길이를 t: Time mask 길이 t0 : Time mask 시작지점 Sampling 방법은 Time Masking 알고리즘과 동일하... DataAugmentationML/DLFeature_DropDataAugmentation