Dataset · Funtion
CrowdHuman Dense Crowd Detection Benchmark
Funtion/CrowdHuman
24,370 images with 470,000 annotated people for evaluating object detectors.
What's inside
A well-known object detection benchmark designed to test pedestrian detection quality in crowds. The dataset is split into training (15,000 photos), validation (4,370 photos), and test (5,000 photos) sets, with 470,000 person instances annotated across the train and validation splits.
What you can do with it
- Training and fine-tuning object detectors (YOLO, DETR) for heavily occluded environments
- Benchmarking computer vision algorithms on the standard CrowdHuman dataset
- Tasks
- Object detection
- Languages
- English
- Size
- 10K–100K rows
- Downloads this month
- 21.3K
- Likes
- 8
How to load it
from datasets import load_dataset
ds = load_dataset("Funtion/CrowdHuman")