论文标题
FBK-HUPBA提交给Epic-Kitchens Action识别2020挑战
FBK-HUPBA Submission to the EPIC-Kitchens Action Recognition 2020 Challenge
论文作者
论文摘要
在本报告中，我们描述了我们提交给Epic-Kitchens Action Denciention 2020挑战的技术细节。为了参与挑战，我们部署了我们最近开发的时空特征提取和聚合模型：GATE-SHIFT模块（GSM）[1]和Egoaco，《长期短期关注》（LSTA）的扩展[2]。我们设计了具有不同骨架和预训练的GSM和Egoaco模型家族的合奏，以产生预测分数。我们的提交内容在公众排行榜上使用Team Name FBK-Hupba可见，在S1设置上获得了40.0％的前1个动作识别精度，仅使用RGB，在S2设置上获得了25.71％的识别精度。
In this report we describe the technical details of our submission to the EPIC-Kitchens Action Recognition 2020 Challenge. To participate in the challenge we deployed spatio-temporal feature extraction and aggregation models we have developed recently: Gate-Shift Module (GSM) [1] and EgoACO, an extension of Long Short-Term Attention (LSTA) [2]. We design an ensemble of GSM and EgoACO model families with different backbones and pre-training to generate the prediction scores. Our submission, visible on the public leaderboard with team name FBK-HUPBA, achieved a top-1 action recognition accuracy of 40.0% on S1 setting, and 25.71% on S2 setting, using only RGB.
