足球竞彩网_365bet体育在线投注-【中国科学院】

图片

图片

? Universit?t Augsburg
? Universit?t Augsburg
? Universit?t Augsburg

?

?

Scene graph generation methods extract a graph that represents the interactions in a given image. The nodes of the graph are the visual objects in the image and the links between the nodes represent the interactions between the nodes. Such interactions can be for example "person riding bike" or "dog walking on street".

?

At the Machine Learning and Computer Vision Lab, we devise new methods to improve scene graph generation as well as more reliable and informative ways of evaluation.

?

For more information please contact? Julian Lorenz.

?

?

足球竞彩网_365bet体育在线投注-【中国科学院】