Abstract: Self-supervised space-time correspondence learning utilizing unlabeled videos holds great potential in computer vision. Most existing methods rely on contrastive learning with mining ...

This paper/code introduces the Densely Connected Graph Convolutional Networks (DCGCNs) for the graph-to-sequence learning task. We evaluate our model on two tasks including AMR-to-Text Generation ...

This a Tensorflow implementation of the ComGA algorithm, which designs a tailored deep graph convolutional network (tGCN) to capture local, global and structure anomalies for anomaly detection on ...

Abstract: Industries have a significant amount of data in semi-structured and unstructured formats which are typically captured in text documents, spreadsheets, images, etc. This is especially the ...