Existing object detection methods are bounded in a fixed-set vocabulary ...
Vision-and-Language Navigation (VLN) is a task that an agent is required...
Thanks to large-scale labeled training data, deep neural networks (DNNs)...
Existing domain adaptation methods on visual sentiment classification
ty...
We study the underlying structure of data (approximately) generated from...