Noisy annotations such as missing annotations and location shifts often exist in crowd counting datasets due to multi-scale head sizes, high occlusion, etc. These noisy annotations severely affect the model training, especially for density map-based methods. To alleviate the negative impact of noisy annotations, we propose a novel crowd counting model with one convolution head and one transformer head, in which these two heads can supervise each other in noisy areas, called Cross-Head Supervision. The resultant model, CHS-Net, can synergize different types of inductive biases for better counting. In addition, we develop a progressive cross-head supervision learning strategy to stabilize the training process and provide more reliable supervision. Extensive experimental results on ShanghaiTech and QNRF datasets demonstrate superior performance over state-of-the-art methods. Code is available at https://github.com/RaccoonDML/CHSNet.
@article{arxiv.2303.09245,
title = {Cross-head Supervision for Crowd Counting with Noisy Annotations},
author = {Mingliang Dai and Zhizhong Huang and Jiaqi Gao and Hongming Shan and Junping Zhang},
journal= {arXiv preprint arXiv:2303.09245},
year = {2023}
}