A critical step in virus infection is the initial uptake of viruses into host cells. Interestingly, and perhaps counterintuitively, a particular virus may use different modes of crossing the plasma membrane, either in the same cell or depending on cell type. Even for members of the same virus family, different modes of uptake have been observed. The main uptake pathways include the use of membrane fusion or endocytosis. A number of endocytic mechanisms have been discovered and these depend on multiple factors, such as cell type and surface receptors, as well as cell membrane properties. Importantly, the uptake path usually determines the fate and infectivity of a virus. Thus, predicting the mode of entry is important for choosing and optimising preventative or impeding measures of infections. However, determining an uptake mechanism is challenging due to multiple factors involving assay design and observation technology. An accurate investigation requires an adapted experimental design for each virus, receptor and host cell, analysing binding, uptake, routing, and the infection cycle of cellular and virus proteins. At the same time, direct microscopic imaging of the viruses and other molecules involved in the uptake into living cells should employ as high a temporal and spatial resolution as possible, since every entry mode displays specific spatiotemporal dynamics of virus and molecular movements at the cell and virus surface. Finally, data analysis needs to be tailored to correlate a “microscopic signature” of entry events with fusion signatures and particular endocytic mechanisms, which would then make it possible to identify the mode(s) of entry of an emerging virus. Unfortunately, the realisation of high-resolution microscope experiments of virus uptake is usually too complex to accomplish a straightforward and quick prediction of entry modes of newly emerging viruses. This is, however, within the overall aim of this CRC VirusREvolution, which is to set up tools to quickly react to potential pandemic threats. We therefore propose to realise a fast and facile tool to predict the mode(s) of virus entry from standard low-resolution live-cell microscopy data. To achieve this, we will combine high spatiotemporal resolution microscopy data with methods from Machine Learning (ML) / Artificial Intelligence (AI) for fast, objective, and quantitative analysis to establish a tool that allows correlation with the standard microscopy data of lower spatiotemporal resolution. This shall enable the characterisation and differentiation of the membrane interactions of virtually any virus, using widely used standard and less complex microscopy approaches, such as widefield or confocal microscopy. We will employ a machine learning approach that builds on the experience gained from working with a variety of viruses. Thus, the performance of our tool will gradually improve as more and more virus data becomes available. Our experimental and computational tools will contribute to answering the research questions of the CRC VirusREvolution, such as revealing mechanisms of virus entry and tackling the goals of the description and prediction of the infectivity of viruses and their hosts.