Publication 04 · Crowd counting

Attention Mixture Network via Binarization Transfer.

Two kinds of attention answer two different questions — where people are, and how dense it is. AMNet refuses to pick one.

Published · ACM MM 2024 (McGE) Crowd counting Dual attention Binarization transfer

Junjie Xu*, Zihao Zhang*, Xin Li, Weijie Li, Kun Yu† · East China Normal University (* equal contribution, † corresponding)

Binary attention maps tell a crowd counter where the crowd is; non-binary maps capture the density gradient inside it. Prior work used one or the other. AMNet feeds shared features to both an attention-map generator and a density-map generator, and couples them through a Binarization Transfer Module (BTM) that jointly optimizes the binarization during training — filtering cluttered backgrounds while keeping density structure.

4datasets: ShanghaiTech A/B, UCF_CC_50, WorldExpo'10, UCF-QNRF
2attention map types exploited simultaneously
1Binarization Transfer Module coupling them
Multi-scalefeature fusion against background noise
01

The question

Crowd counting lives or dies on attention. The binary kind separates crowd regions from intricate background; the non-binary kind characterizes congestion level inside those regions. Methods historically commit to one — and pay for it either in background noise or in lost density structure.

The question: can both maps be exploited at once, with the binarization itself learned rather than hand-thresholded?

02

The approach

AMNet shares a feature extractor between an attention map generator (binary, candidate crowd regions) and a density map generator (non-binary, congestion gradients). The Binarization Transfer Module turns binarization into a differentiable, jointly-optimized step — transferring the clean region selection from the binary map into the density estimation while preserving multi-scale detail.

Traditional attention-based crowd counting pipeline versus the AMN pipeline with binary and density attention maps
Fig. 1Traditional attention-based pipelines (blue and yellow flows) vs the AMN pipeline (red flow) — binary and density attention maps cooperate instead of competing.
03

What it shows

Across ShanghaiTech PartA and PartB, UCF_CC_50, WorldExpo'10, and UCF-QNRF, AMNet achieves significant improvements in counting accuracy and outperforms state-of-the-art methods — with the largest gains exactly where binary attention alone fails: dense, cluttered scenes.

Qualitative density maps show the mechanism working as designed: backgrounds suppressed by the binarized attention while density gradients inside the crowd remain intact.

AMNet architecture: shared features feeding attention map generator and density map generator with multi-scale fusion
Fig. 2The AMNet architecture — feature maps flow to the attention map generator (AMG) and density map generator (DMG), fused multi-scale for the final density estimate.
04

Claim boundary

✓ Established

  • Simultaneous use of binary and non-binary attention maps
  • A differentiable Binarization Transfer Module for joint training
  • Improvements across four standard crowd-counting benchmarks
  • Qualitative evidence of background suppression with preserved density

✗ Explicitly not claimed

  • Benchmark scope: the five standard counting sets tested
  • No deployment or real-time claim
  • Privacy context of crowd imagery acknowledged, not studied
  • No cross-domain (non-crowd) transfer claim
05

Citation

@inproceedings{xu2024amnet,
  title     = {Attention Mixture Network for Crowd Counting
               via Binarization Transfer},
  author    = {Xu, Junjie and Zhang, Zihao and Li, Xin and Li, Weijie
               and Yu, Kun},
  booktitle = {Proceedings of the 2nd International Workshop on Multimedia
               Content Generation and Evaluation (McGE)}, 
  series    = {ACM Multimedia 2024},
  year      = {2024},
  doi       = {10.1145/3688867.3690172}
}