MLP-based weakly supervised crowd counting approaches have made significant advancements over the past few years. However, owing to the limited datasets, the current MLP-based methods do not consider the problem of region-to-region dependency in the image. For this, we propose a weakly supervised method termed SR2. SR2 consists of three parts: scale-reasoning module, scale-ranking module, and regression branch. In particular, the scale-reasoning module extracts and fuses the region-to-region dependency in the image and multiple scale feature, then sends the fused features to the regression branch to obtain estimated counts; the scale-ranking module is used to understand the internal information of the image better and expand the datasets efficiently, which will help to improve the accuracy of the estimated counts in the regression branch. We conducted extensive experiments on four benchmark datasets. The final results showed that our approach has better and higher competing counting performance with respect to other weakly supervised counting networks and with respect to some popular fully supervised counting networks.

The image edge is the first glance to observe the external features of things. Face edge extracting in face recognition is of great significance. The face edge contains important face information, which not only filters out the interference information, but also retains useful information, so that the calculation time of the computer will be reduced to a certain extent. This paper proposes an algorithm with which the face edge is extracted by a kind of operator (i.e. LoG operator) and the face recognition is completed by BP neural network. Many former experiments showed that the LoG operator has some limitations in the extraction of face edge. To solve the limitations of the LoG operator, we adjust the weights of the Gaussian filter in the LoG operator and define where the face edge is by gray level co-occurrence matrix. The experiments show that the improved LoG operator achieves better results compared to the traditional one.