Boosting Detection in Crowd Analysis via Underutilized Output Features
Authors:
Preprint
Abstract
Detection-based methods have been viewed unfavorably in crowd analysis due to their poor performance in dense crowds. However, we argue that the potential of these methods has been underestimated, as they offer crucial information for crowd analysis that is often ignored. Specifically, the area size and confidence score of output proposals and bounding boxes provide insight into the scale and density of the crowd. To leverage these underutilized features, we propose Crowd Hat, a plug-and-play module that can be easily integrated with existing detection models. This module uses a mixed 2D-1D compression technique to refine the output features and obtain the spatial and numerical distribution of crowd-specific information. Based on these features, we further propose region-adaptive NMS thresholds and a decouple-then-align paradigm that address the major limitations of detection-based methods. Our extensive evaluations on various crowd analysis tasks, including crowd counting, localization, and detection, demonstrate the effectiveness of utilizing output features and the potential of detection-based methods in crowd analysis.Keywords
All Related Versions
This publication has 9 references indexed in Scilit:
- Interpretable Sparsification of Brain Graphs: Better Practices and Effective Designs for Graph Neural Networks2023
- Gradient-less Federated Gradient Boosting Tree with Learnable Learning Rates2023
- An End-to-End Transformer Model for Crowd Localization2022
- CrowdFormer: An Overlap Patching Vision Transformer for Top-Down Crowd Counting2022
- AutoScale: Learning to Scale for Crowd CountingInternational Journal of Computer Vision, 2022
- A Self-Training Approach for Point-Supervised Object Detection and Counting in CrowdsIEEE Transactions on Image Processing, 2021
- NWPU-Crowd: A Large-Scale Benchmark for Crowd Counting and LocalizationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2020
- Nuclear Norm Based Matrix Regression with Applications to Face Recognition with Occlusion and Illumination ChangesIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016
- Cumulative Attribute Space for Age and Crowd Density Estimation2013