crowd-counting
The dataset comprises 647 crowd images, with up to 11,000 individuals, and provides keypoint annotations for precise crowd counting and density estimation. It is designed for crowd counting tasks, especially in highly congested scenes, addressing challenges of varying scales and density. The dataset includes both dense and sparse crowd examples, offering density maps for estimation and supporting crowd monitoring and object detection applications. It enables deep‑learning models to improve crowd control and management, providing rich data such as feature maps and predicted density outputs, and aids public safety monitoring in venues with diverse crowd densities.
Dataset description and usage context
Crowd Density Dataset - Different Crowd Sizes
Dataset Overview
- Number of Images: 647
- Maximum Individuals: Up to 11,000
- Annotation Type: Keypoints
- Application Scenarios: Crowd counting and density estimation
- Data Characteristics: Includes examples of both dense and sparse crowds for realistic analysis
Intended Uses
- Task Categories:
- Image Classification
- Image‑to‑Image Translation
- Feature Extraction
- Object Detection
- Labels:
- Crowd Counting
- Security
- Public Safety
- Keypoint Detection
- Keypoints
- Annotation
- Crowd
- Public Surveillance
- Crowd Density
- Density Ranges:
- 0‑1,000
- 1,000‑2,000
- 2,000‑3,000
- 3,000‑4,000
- 4,000‑5,000
Advantages
- Density Maps: Provided for density estimation, suitable for monitoring and detection tasks
- Deep‑Learning Support: Enables advanced crowd control and management models
- Rich Features: Includes feature maps and predicted density outputs
Access
- Dataset Access: For the full dataset, contact UniData to discuss requirements and pricing.
Pair the dataset with AI analysis and content workflows.
Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.