Conditional object-centric slot attention learning for video and other sequential data

The present application relates to methods, systems, and non-transitory computer readable media associated with conditional object-centric slot attention learning for video and other sequential data. A method includes obtaining a first feature vector and a second feature vector representing content of first and second image frames of an input video, respectively. The method may also include generating first slot vectors based on the first feature vectors, wherein each slot vector represents an attribute of a corresponding entity as represented in the first image frame; and generating prediction slot vectors based on the first slot vectors, the prediction slot vectors including corresponding prediction slot vectors representing transitions of attributes of corresponding entities from the first image frame to the second image frame. The method may also include generating second slot vectors based on the predicted slot vector and the second feature vector, the second slot vectors including corresponding slot vec