Expert Training: Task Hardness Aware Meta-Learning for Few-Shot Classification release_34qfap2as5bupemk6zdksje3oe

by Yucan Zhou, Yu Wang, Jianfei Cai, Yu Zhou, Qinghua Hu, Weiping Wang

Released as a article .

2020  

Abstract

Deep neural networks are highly effective when a large number of labeled samples are available but fail with few-shot classification tasks. Recently, meta-learning methods have received much attention, which train a meta-learner on massive additional tasks to gain the knowledge to instruct the few-shot classification. Usually, the training tasks are randomly sampled and performed indiscriminately, often making the meta-learner stuck into a bad local optimum. Some works in the optimization of deep neural networks have shown that a better arrangement of training data can make the classifier converge faster and perform better. Inspired by this idea, we propose an easy-to-hard expert meta-training strategy to arrange the training tasks properly, where easy tasks are preferred in the first phase, then, hard tasks are emphasized in the second phase. A task hardness aware module is designed and integrated into the training procedure to estimate the hardness of a task based on the distinguishability of its categories. In addition, we explore multiple hardness measurements including the semantic relation, the pairwise Euclidean distance, the Hausdorff distance, and the Hilbert-Schmidt independence criterion. Experimental results on the miniImageNet and tieredImageNetSketch datasets show that the meta-learners can obtain better results with our expert training strategy.
In text/plain format

Archived Files and Locations

application/pdf  1.7 MB
file_s2nuppnyjrezhcf6bltp7j7z6a
arxiv.org (repository)
web.archive.org (webarchive)
Read Archived PDF
Preserved and Accessible
Type  article
Stage   submitted
Date   2020-07-13
Version   v1
Language   en ?
arXiv  2007.06240v1
Work Entity
access all versions, variants, and formats of this works (eg, pre-prints)
Catalog Record
Revision: d9a86873-32fe-4a0b-9a29-a471095835d9
API URL: JSON