Two-stage architectural fine-tuning with neural architecture search using early-stopping in image classification. (arXiv:2202.08604v3 [cs.CV] UPDATED)

In many deep neural network (DNN) applications, the difficulty of gathering
high-quality data in the industry field hinders the practical use of DNN. Thus,
the concept of transfer learning has emerged, which leverages the pretrained
knowledge of DNNs trained on large-scale datasets. Therefore, this paper
suggests two-stage architectural fine-tuning, inspired by neural architecture
search (NAS). One of main ideas is mutation, which reduces the search cost
using given architectural information. Moreover, early-stopping is considered
which cuts NAS costs by terminating …

architecture arxiv classification early-stopping fine-tuning image neural architecture search search stage