HARD-Net: Hardness-AwaRe Discrimination Network for 3D Early Activity Prediction

Abstract

Predicting the class label from the partially observed activity sequence is a very hard task, as the observed early segments of different activities can be very similar. In this paper, we propose a novel Hardness-AwaRe Discrimination Network (HARD-Net) to specifically investigate the relationships between the similar activity pairs that are hard to be discriminated. Specifically, a Hard Instance-Interference Class (HI-IC) bank is designed, which dynamically records the hard similar pairs. Based on the HI-IC bank, a novel adversarial learning scheme is proposed to train our HARD-Net, which thus grants our network with the strong capability in mining subtle discrimination information for 3D early activity prediction. We evaluate our proposed HARD-Net on two public activity datasets and achieve state-of-the-art performance.

Cite

Text

Li et al. "HARD-Net: Hardness-AwaRe Discrimination Network for 3D Early Activity Prediction." Proceedings of the European Conference on Computer Vision (ECCV), 2020. doi:10.1007/978-3-030-58621-8_25

Markdown

[Li et al. "HARD-Net: Hardness-AwaRe Discrimination Network for 3D Early Activity Prediction." Proceedings of the European Conference on Computer Vision (ECCV), 2020.](https://mlanthology.org/eccv/2020/li2020eccv-hardnet/) doi:10.1007/978-3-030-58621-8_25

BibTeX

@inproceedings{li2020eccv-hardnet,
  title     = {{HARD-Net: Hardness-AwaRe Discrimination Network for 3D Early Activity Prediction}},
  author    = {Li, Tianjiao and Liu, Jun and Zhang, Wei and Duan, Lingyu},
  booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
  year      = {2020},
  doi       = {10.1007/978-3-030-58621-8_25},
  url       = {https://mlanthology.org/eccv/2020/li2020eccv-hardnet/}
}