Mask Focal Loss: A unifying framework for dense crowd counting with canonical object detection networks

TL;DR

This paper introduces Mask Focal Loss, a novel loss function for dense crowd counting that improves detection accuracy by addressing sample imbalance and spatial coherence issues, validated on a new synthetic dataset.

Contribution

The paper proposes Mask Focal Loss, a unifying loss framework for heatmap-based crowd counting, and introduces GTA_Head, a synthetic dataset for evaluation.

Findings

Reduces MAE by up to 47.03%

Reduces RMSE by up to 61.99%

Demonstrates superior performance across detectors and datasets

Abstract

As a fundamental computer vision task, crowd counting plays an important role in public safety. Currently, deep learning based head detection is a promising method for crowd counting. However, the highly concerned object detection networks cannot be well applied to this problem for three reasons: (1) Existing loss functions fail to address sample imbalance in highly dense and complex scenes; (2) Canonical object detectors lack spatial coherence in loss calculation, disregarding the relationship between object location and background region; (3) Most of the head detection datasets are only annotated with the center points, i.e. without bounding boxes. To overcome these issues, we propose a novel Mask Focal Loss (MFL) based on heatmap via the Gaussian kernel. MFL provides a unifying framework for the loss functions based on both heatmap and binary feature map ground truths. Additionally, we introduce GTA_Head, a synthetic dataset with comprehensive annotations, for evaluation and comparison. Extensive experimental results demonstrate the superior performance of our MFL across various detectors and datasets, and it can reduce MAE and RMSE by up to 47.03% and 61.99%, respectively. Therefore, our work presents a strong foundation for advancing crowd counting methods based on density estimation.