TLDChoiceNet: Quantitatively Choosing a Transfer Learning Dataset

Jing Ning, James D. Braza

Transfer learning is particularly useful in settings with limited training data, and within image classification it is common to transfer learn upon massive datasets like ImageNet…