Uploaded March 2024 | Updated September 2026, 1 week ago
Learn how to perform data augmentation with KerasCV. Wei, a Developer Advocate at Google, covers how to augment image data with some of the most popular and useful augmentation layers: ‘RandAugment,’ ‘CutMix,’ and ‘MixUp.’ These layers are used in nearly all state-of-the-art image classification pipelines.
Chapters:
0:00 - Introduction
0:56 - Layers
2:24 - Customizing augmentation pipeline
3:12 - Training a CNN with augmentation
Resources:
Tutorials
CutMix, MixUp, and RandAugment image augmentation with KerasCV → https://goo.gle/3T6iS0Q
Classification with KerasCV → https://goo.gle/3wDueSo
Custom Image Augmentations with BaseImageAugmentationLayer → https://goo.gle/3Taey0w
Papers
CutMix: Train Strong Classifiers Paper → https://goo.gle/3IrW2Md
MixUp: Beyond Empirical Risk Minimization Paper → https://goo.gle/3It4pHm
Watch more Applied ML with KerasCV and KerasNLP → https://goo.gle/AppliedMLwithKeras
Subscribe to the TensorFlow channel → https://goo.gle/TensorFlow
#TensorFlow
Speaker: Wei Wei
Learn how to perform data augmentation with KerasCV. Wei, a Developer Advocate at Google, covers how to augment image data with some of the most popular and useful augmentation layers: ‘RandAugment,’ ‘CutMix,’ and ‘MixUp.’ These layers are used in nearly all state-of-the-art image classification pipelines.
Chapters:
0:00 - Introduction
0:56 - Layers
2:24 - Customizing augmentation pipeline
3:12 - Training a CNN with augmentation
Resources:
Tutorials
CutMix, MixUp, and RandAugment image augmentation with KerasCV → https://goo.gle/3T6iS0Q
Classification with KerasCV → https://goo.gle/3wDueSo
Custom Image Augmentations with BaseImageAugmentationLayer → https://goo.gle/3Taey0w
Papers
CutMix: Train Strong Classifiers Paper → https://goo.gle/3IrW2Md
MixUp: Beyond Empirical Risk Minimization Paper → https://goo.gle/3It4pHm
Watch more Applied ML with KerasCV and KerasNLP → https://goo.gle/AppliedMLwithKeras
Subscribe to the TensorFlow channel → https://goo.gle/TensorFlow
#TensorFlow
Speaker: Wei Wei










