Abstract | ||
---|---|---|
•A scheduled sampling strategy is introduced to adjust the training procedure of matching network, which accomplishes to learn the ability for one-shot prediction from easy to difficult.•We propose a novel metric to measure the difficulty of training samples, which jointly considers the diversity and similarity among the labels’ semantic. Samples with high-difficulty values are more difficult to learn for the matching network.•We conduct extensive experiments on datasets mini-Imagenet, Birds, and Flowers to illustrate the effectiveness and superiority of the proposed method. The experimental results demonstrate that our method consistently outperforms other competitors. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1016/j.patcog.2019.07.007 | Pattern Recognition |
Keywords | Field | DocType |
Scheduled sampling,Matching network,From easy to difficult,One-shot learning,Difficulty metric | Pattern recognition,Artificial intelligence,Sampling (statistics),One-shot learning,Learning disability,Machine learning,Mathematics | Journal |
Volume | Issue | ISSN |
96 | 1 | 0031-3203 |
Citations | PageRank | References |
2 | 0.36 | 0 |
Authors | ||
6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Lingling Zhang | 1 | 276 | 45.79 |
Jun Liu | 2 | 178 | 25.96 |
Minnan Luo | 3 | 269 | 21.18 |
Xiaojun Chang | 4 | 1585 | 76.85 |
Qinghua Zheng | 5 | 1261 | 160.88 |
Alexander G. Hauptmann | 6 | 7472 | 558.23 |