Coarsely-labeled Data for Better Few-shot Transfer

Cheng Perng Phoo, Bharath HariharanICCV 2021 · OpenReview

Abstract

Few-shot learning is based on the premise that labels are expensive, especially when they are fine-grained and require expertise. But coarse labels might be easy to acquire and thus abundant. We present a representation learning approach - PAS that allows few-shot learners to leverage coarsely-labeled data available before evaluation. Inspired by self-training, we label the additional data using a teacher trained on the base dataset and filter the teacher’s prediction based on the coarse labels; a new student representation is then trained on the base dataset and the pseudo-labeled dataset. PAS is able to produce a representation that consistently and significantly outperforms the baselines in 3 different datasets. Code is available at https://github.com/cpphoo/PAS
BibTeX
@inproceedings{phoo2021coarsely,
  author = {Cheng Perng Phoo and Bharath Hariharan},
  title = {{Coarsely-labeled Data for Better Few-shot Transfer}},
  booktitle = {IEEE/CVF International Conference on Computer Vision},
  year = {2021},
  url = {https://openreview.net/forum?id=_35h8iHkU2},
}

Loading related content...

Figure