Scene-specific crowd counting using synthetic training images

Crowd counting is a computer vision task on which considerable progress has recently been made thanks to convolutional neural networks. However, it remains challenging even in scene-specific settings, real-world application scenarios where no representative images of the target scene are available, not unlabelled, for training or fine-tuning crowd model. Inspired by previous work other tasks, we propose simple but effective solution above scenario, consists automatically building set synthetic images. Our does require from end-users any manual annotation effort nor collection scene. Extensive experiments several benchmark data sets show that proposed can improve effectiveness existing methods.