Due to the ensemble effects of label of each pixel, bounding boxes inferred from semantic segmentation map are more stable and accurate than detection methods. This paper uses semantic segmentation map to find the bounding boxes of human body parts. U_Net [12] is a popular semantic segmentation...
Based on bag of words (BoW) model, topic models are used to cluster a set of discriminative features for classes of interest. However, features in topic models have no spatial information, thus, it cannot explicitly model the intra- and inter-class structure of feature distributions. Recently,...