Convolutional Long Short-Term Memory Networks for Recognizing First Person Interactions

We present a novel deep learning approach for addressing the problem of interaction recognition from a first person perspective. The approach uses a pair of convolutional neural networks, whose parameters are shared, for extracting frame level features from successive frames of the video. The frame level features are then aggregated using a convolutional long short-term memory. The final hidden state of the convolutional long short-term memory is used for classification in to the respective categories. In our network the spatio-temporal structure of the input is preserved till the very final processing stage. Experimental results show that our method outperforms the state of the art on most recent first person interactions datasets that involve complex ego-motion. On UTKinect, it competes with methods that use depth image and skeletal joints information along with RGB images, while it surpasses previous methods that use only RGB images by more than 20% in recognition accuracy.