Image Fusion Transformer

Over a year has passed since the finalization of VersatileVideo Coding (H.266/VVC), yet it is still far from practicaldeployment, a major reason being the excessive complexity. The flexible and sophisticated quad-tree with nested multi-type tree partitioning structure in VVC provides considerableperformance gains while bringing about an exponential in-crease in encoding time. To reduce the coding complexity,this paper proposes a Convolutional Neural Network (CNN) based fast Coding Unit (CU) partitioning algorithm for intracoding, which accelerates CU partition through predicting thepartition modes with texture information and terminating redundant modes in advance. Corresponding classifiers are designed for different CU sizes to improve prediction accuracy. Low rate-distortion performance degradation is guaranteed byintroducing performance loss due to misclassification into theloss function. Experiments show that the proposed methodcan save encoding time ranging from 38.39% to 62.33% with 0.92% to 2.36% bit rate increase.