Domain-adaptive Crowd Counting via Inter-domain Features Segregation and Gaussian-prior Reconstruction

Recently, crowd counting using supervised learning achieves a remarkable improvement. Nevertheless, most counters rely on a large amount of manually labeled data. With the release of synthetic crowd data, a potential alternative is transferring knowledge from them to real data without any manual label. However, there is no method to effectively suppress domain gaps and output elaborate density maps during the transferring. To remedy the above problems, this paper proposed a Domain-Adaptive Crowd Counting (DACC) framework, which consists of Inter-domain Features Segregation (IFS) and Gaussian-prior Reconstruction (GPR). To be specific, IFS translates synthetic data to realistic images, which contains domain-shared features extraction and domain-independent features decoration. Then a coarse counter is trained on translated data and applied to the real world. Moreover, according to the coarse predictions, GPR generates pseudo labels to improve the prediction quality of the real data. Next, we retrain a final counter using these pseudo labels. Adaptation experiments on six real-world datasets demonstrate that the proposed method outperforms the state-of-the-art methods. Furthermore, the code and pre-trained models will be released as soon as possible.