Learning Error-Driven Curriculum for Crowd Counting
This work addresses a specific bottleneck in crowd counting for applications like surveillance and event management, representing an incremental improvement.
The paper tackles the problem of frequency imbalance in density maps for crowd counting by proposing an error-driven curriculum learning strategy, achieving state-of-the-art performance on two benchmark datasets.
Density regression has been widely employed in crowd counting. However, the frequency imbalance of pixel values in the density map is still an obstacle to improve the performance. In this paper, we propose a novel learning strategy for learning error-driven curriculum, which uses an additional network to supervise the training of the main network. A tutoring network called TutorNet is proposed to repetitively indicate the critical errors of the main network. TutorNet generates pixel-level weights to formulate the curriculum for the main network during training, so that the main network will assign a higher weight to those hard examples than easy examples. Furthermore, we scale the density map by a factor to enlarge the distance among inter-examples, which is well known to improve the performance. Extensive experiments on two challenging benchmark datasets show that our method has achieved state-of-the-art performance.