Absolute Zero-Shot Learning

Abstract

Considering the increasing concerns about data copyright and privacy issues, we present a novel Absolute Zero-Shot Learning (AZSL) paradigm, i.e., training a classifier with zero real data. The key innovation is to involve a teacher model as the data safeguard to guide the AZSL model training without data leaking. The AZSL model consists of a generator and student network, which can achieve date-free knowledge transfer while maintaining the performance of the teacher network. We investigate black-box' and white-box’ scenarios in AZSL task as different levels of model security. Besides, we also provide discussion of teacher model in both inductive and transductive settings. Despite embarrassingly simple implementations and data-missing disadvantages, our AZSL framework can retain state-of-the-art ZSL and GZSL performance under the white-box' scenario. Extensive qualitative and quantitative analysis also demonstrates promising results when deploying the model under black-box’ scenario.

@misc{gao2022absolute,
      title={Absolute Zero-Shot Learning}, 
      author={Rui Gao and Fan Wan and Daniel Organisciak and Jiyao Pu and Junyan Wang and Haoran Duan and Peng Zhang and Xingsong Hou and Yang Long},
      year={2022},
      eprint={2202.11319},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}