QD<sup>2</sup>CrowdNet: A Quality Enhanced Depthwise-Dilated Architecture for Efficient and High-Fidelity Crowd Density Estimation
<p dir="ltr">Crowd counting and density estimation play a vital role in public safety, intelligent transport, urban planning, and event management. However, challenges like occlusions, complex scenes, and scale variation make accurate estimation difficult.This project introduces QD²C...
Saved in:
| Main Author: | |
|---|---|
| Published: |
2025
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | <p dir="ltr">Crowd counting and density estimation play a vital role in public safety, intelligent transport, urban planning, and event management. However, challenges like occlusions, complex scenes, and scale variation make accurate estimation difficult.This project introduces QD²CrowdNet, a lightweight yet effective deep learning model that balances accuracy and efficiency. It uses depthwise separable convolutions, a dilated multiscale backend, and a refinement module to generate high-quality density maps, making it ideal for real-time and edge-based applications.</p> |
|---|