Y Su, F Jurie - International Journal of Computer Vision, 2012 - hal.science
The Bag-of-Words (BoW) model-commonly used for image classification-has two strong
limitations: on one hand, visual words lack semantic meanings, on the other hand, they are …