74K views · 147 shares | Microsoft Vision Model ResNet-50 is a state-of-the-art pretrained vision model that uses multi-task learning and web supervision for high performance with lower cost in production settings. In our webinar, learn to use the model to build an image classifier: https://aka.ms/AAb5uto | Microsoft Research | Facebook

44K views · 72 shares | Modern vision systems are highly sensitive to perturbations of their inputs. In new work, researchers present a framework that exploits this sensitivity to create “unadversarial objects,” inputs optimized specifically for more robust model performance: https://aka.ms/AAao4mz | Microsoft Research | Facebook

22K views · 142 reactions | In vision-and-language pretraining (VLP), objects can be used as anchor points to make aligning semantics between image-text pairs easier. Learn how Oscar, a novel VLP framework utilizing objects, sets new state of the art on six vision-and-language tasks: https://aka.ms/AA8flix | Microsoft Research | Facebook