We propose a framework for early action recognition and anticipation by
correlating past features with the future using three novel similarity measures
called Jaccard vector similarity, Jaccard cross-correlation and Jaccard
Frobenius inner product over covariances. Using these combinations of novel
losses and using our framework, we obtain state-of-the-art results for early
action recognition in UCF101 and JHMDB datasets by obtaining 91.7 % and 83.5 %
accuracy respectively for an observation percentage of 20. Similarly, we obtain
state-of-the-art results for Epic-Kitchen55 and Breakfast datasets for action
anticipation by obtaining 20.35 and 41.8 top-1 accuracy respectively.

提出一种早期行动识别和预测的框架，使用三种称为 Jaccard 向量相似性、Jaccard 交叉相关和 Jaccard Frobenius 内积的新相似性度量来关联过去的特征和未来的特征，利用这些创新的损失和我们的框架，在 UCF101 和 JHMDB 数据集中获得了 91.7％和 83.5％的准确性，在观察百分比 20 的情况下，类似地，我们在 Epic-Kitchen55 和 Breakfast 数据集中获得了 20.35 和 41.8 的 top-1 准确度，取得了最先进的结果。