Action Recognition Using Multiple Pooling Strategies of CNN Features
Action Recognition Using Multiple Pooling Strategies of CNN Features is a scholarly work, published in 2018 in ''Neural Processing Letters''. The main subjects of the publication include political representation, pooling, anomaly detection, convolutional neural network, convolution, artificial intelligence, feature extraction, pyramid, discriminative model, feature, Multiple object tracking, pattern recognition, and computer science. The authors explore three different pooling strategies, which are called space-time feature pooling (STFP), time filter pooling (TFP) and spatio-temporal pyramid pooling (STPP), respectively.