Deep learning models for crowd counting in images
Hossain, Mohammad Asiful
Crowd counting on images has become a challenging task for computer vision research. Given an image of a crowded scene, our goal is to estimate the density map of this image, where each pixel value in the density map corresponds to the crowd density at the corresponding location in the image. Given the estimated density map, the final crowd count can be obtained by summing over all values in the density map. One challenge of crowd counting is the scale variation in images. In this thesis, we propose different deep learning models to solve problems regarding crowd counting. This thesis consists of three works which are disjoint but problem domain is similar which is crowd counting. We got reasonably better performance in all these works on benchmark dataset.