CHIRLA: Comprehensive High-resolution Identification and Re-identification for Large-scale Analysis

🏆 Leaderboard: Re-ID and tracking results, sortable, with how to submit · 🌐 Project page

The CHIRLA dataset (Comprehensive High-resolution Identification and Re-identification for Large-scale Analysis) is designed for long-term person re-identification (Re-ID) in real-world scenarios. The dataset consists of multi-camera video recordings captured over seven months in an indoor office environment, featuring 22 individuals with 963,554 bounding box annotations across 596,345 frames.

This dataset aims to facilitate the development and evaluation of Re-ID algorithms capable of handling significant variations in individuals' appearances, including changes in clothing and physical characteristics over extended time periods.

📊 Dataset Statistics

Metric	Value
Duration	7 months
Individuals	22 unique persons
Cameras	7 multi-view cameras
Video Files	70 sequences
Total Frames	596,345 frames
Annotations	963,554 bounding boxes
Resolution	1080×720 pixels
Frame Rate	30 fps
Environment	Indoor office setting

🎥 Data Generation Procedures

The dataset was recorded at the Robotics, Vision, and Intelligent Systems Research Group headquarters at the University of Alicante, Spain. Seven strategically placed Reolink RLC-410W cameras were used to capture videos in a typical office setting, covering areas such as laboratories, hallways, and shared workspaces. Each camera features a 1/2.7" CMOS image sensor with a 5.0-megapixel resolution and an 80° horizontal field of view. The cameras were connected via Ethernet and WiFi to ensure stable streaming and synchronization.

A ROS-based interconnection framework was used to synchronize and retrieve images from all cameras. The dataset includes video recordings at a resolution of 1080×720 pixels, with a consistent frame rate of 30 fps, stored in AVI format with DivX MPEG-4 encoding.

Data Processing Methods and Steps

Data processing involved a semi-automatic labeling procedure:

1. Automated Detection and Tracking

• Detection: YOLOv8x was used to detect individuals in video frames and extract bounding boxes
• Tracking: The Deep SORT algorithm was employed to generate tracklets and assign unique IDs to detected individuals

2. Manual Verification and Correction

• Custom GUI: A specialized graphical user interface was developed for manual verification and correction
• Identity Consistency: Bounding boxes and IDs were manually verified for consistency across different cameras and sequences
• Quality Control: All annotations underwent thorough manual review to ensure accuracy

🔗 Labeling Tool: The custom GUI used for annotation is available at: CHIRLA Labeling Tool

📁 Data Structure and Format

The dataset comprises:

• Video Files: 70 videos, each corresponding to a specific camera view in a sequence, stored in AVI format
• Annotation Files: JSON files containing frame-wise annotations, including bounding box coordinates and identity labels
• Benchmark Data: Processed image crops organized for ReID and tracking evaluation

Directory Structure

CHIRLA/
├── videos/                          # Raw video files
│   └── seq_XXX/
│       └── camera_Y.avi             # Video files for each camera view
├── annotations/                     # Frame-level annotations
│   └── seq_XXX/
│       └── camera_Y.json            # Bounding boxes and IDs
└── benchmark/                       # Processed benchmark data
    ├── reid/                        # Person Re-Identification
    │   ├── long_term/               # Long-term ReID scenario
    │   │   ├── train/
    │   │   │   ├── train_0/
    │   │   │   │   └── seq_XXX/
    │   │   │   └── train_1/
    │   │   └── test/
    │   │       ├── test_0/          # Validation subset
    │   │       └── test_1/          # Test subset
    │   ├── multi_camera/            # Multi-camera ReID
    │   ├── multi_camera_long_term/  # Combined scenario
    │   └── reappearance/            # Reappearance detection
    └── tracking/                    # Person Tracking
        ├── brief_occlusions/        # Short-term occlusions
        └── multiple_people_occlusions/  # Multi-person scenarios

📥 CHIRLA Dataset Downloader

The CHIRLA dataset is hosted on 🤗 Hugging Face Datasets:
👉 bdager/CHIRLA

We also provide a CLI downloader for direct access.

For detailed usage instructions (both Hugging Face and CLI options), see downloader/README.md

📈 Benchmark

We propose a benchmark with different scenarios for tracking and reidentification tasks. Please see benchmark/README.md to have all the information on how to run your methods on the different challenges of the benchmark.

Use Cases and Reusability

The CHIRLA dataset is suitable for:

• Long-term person re-identification
• Multi-camera tracking and re-identification
• Single-camera tracking and re-identification

License

The dataset is publicly available under the CC-BY license. The code in this repository is released under the MIT License.

Citation

If you use CHIRLA dataset and benchmark, please cite the work as:

@article{bdager2025chirla, 
    title={CHIRLA: Comprehensive High-resolution Identification and Re-identification for Large-scale Analysis}, 
    author={Dominguez-Dager, Bessie and Escalona, Felix and Gomez-Donoso, Fran and Cazorla, Miguel},  
    journal={arXiv preprint arXiv:2502.06681}, 
    year={2025}, 
}

Contact

For any questions or support, please contact bessie.dominguez@ua.es