爱奇艺提交ActivityNet Challenge 2019 Kinetics-700 challenge：分层分组关注
iqiyi Submission to ActivityNet Challenge 2019 Kinetics-700 challenge: Hierarchical Group-wise Attention
摘要：
在这份报告中，爱奇艺提交ActivityNet 2019 Kinetics-700挑战任务的方法被描述。模式集成阶段涉及三个模式：TSN，HG-NL和StNet。我们提出了分层组明智的非本地（HG-NL）模块的帧级特征聚合的视频分类。标准的非局部（NL）模块在视频分类任务中聚集帧级特征是有效的，但存在参数效率低和计算成本高。HG-NL方法涉及分层分组结构并生成多个注意力地图以增强性能。基于这种分层的分组结构，该方法具有竞争力的精度，更少的参数和更小的计算成本比标准NL。对于ActivityNet 2019 Kinetics-700挑战的任务，在模型集成之后，我们最终在测试集上获得了平均前1和前5错误百分比28.444%。
英文摘要：
In this report, the method for the iqiyi submission to the task of ActivityNet 2019 Kinetics-700 challenge is described. Three models are involved in the model ensemble stage: TSN, HG-NL and StNet. We propose the hierarchical group-wise non-local (HG-NL) module for frame-level features aggregation for video classification. The standard non-local (NL) module is effective in aggregating frame-level features on the task of video classification but presents low parameters efficiency and high computational cost. The HG-NL method involves a hierarchical group-wise structure and generates multiple attention maps to enhance performance. Basing on this hierarchical group-wise structure, the proposed method has competitive accuracy, fewer parameters and smaller computational cost than the standard NL. For the task of ActivityNet 2019 Kinetics-700 challenge, after model ensemble, we finally obtain an averaged top-1 and top-5 error percentage 28.444% on the test set.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
