Special issue on deep learning in image and video retrieval

Ard Oerlemans, Yanming Guo, Michael S. Lew, Tat-Seng Chua
2020 International Journal of Multimedia Information Retrieval  
In recent years, deep learning techniques have been rapidly evolving and advancing, which has resulted in such approaches finding their way into almost all fields where machine learning had already been used or where classic algorithms were still the chosen solution. From image classification and image segmentation to natural language translation and climate modeling, the deep learning approaches have consistently outperformed the best previous algorithms. This special issue aims to capture the
more » ... state of the art in deep learning in the context of image and video retrieval. We are pleased to present the following five papers, that were selected after a triple peer-review process: A useful feature in an image retrieval context is the number of people present in an image, especially when it involves a large number of people. The paper "Single Image Crowd Counting: A Comparative Survey on Deep Learning-based Approaches" by V. Nguyen and T.D. Ngo provides an overview of recent research on crowd counting based on single image input. They start with discussing the approaches that were used before the deep learning era and then show the different aspects of recent research in crowd counting, like multi-scale approaches, divide-and-conquer techniques and coarse counting.
doi:10.1007/s13735-020-00194-y fatcat:5c75uxy2cbfavilizrmmtv5qra