Back to datasets
Dataset assetOpen Source CommunityPerson Re-identificationCross-Scenario Surveillance
RAiD_Dataset
The RAiD dataset is an indoor–outdoor cross‑scenario re‑identification dataset collected at the Winstun Chung Hall of UC Riverside. It contains images from four cameras (two indoor and two outdoor), recording 6,920 images of 43 individuals. Detailed annotations include images, foreground masks, camera IDs, and person IDs.
Source
github
Created
Nov 26, 2014
Updated
Feb 23, 2024
Signals
103 views
Availability
Linked source ready
Overview
Dataset description and usage context
RAiD_Dataset Overview
Basic Information
- Name: RAiD_Dataset
- Source: Winstun Chung Hall, UC Riverside
- Type: 4‑camera dataset (2 indoor, 2 outdoor)
- Camera IDs: 1 & 2 are indoor, 3 & 4 are outdoor
- Subjects: 43 people
- Total Images: 6,920
- Special Cases: 41 subjects appear in all 4 cameras; subject 8 absent from camera 3; subject 34 absent from camera 4
Content
- Image Data:
- Format: 4‑D array, 128×64×3×6920
- Content: All detected RGB images, resized to 128×64
- Mask Data:
- Format: 3‑D binary array, 128×64×6920
- Content: Foreground masks for the images
- Camera Assignment Info:
- Format: Vector of length 6920
- Content: Camera ID for each detection
- Person ID Info:
- Format: Vector of length 6920
- Content: Person ID for each detection
- Detection Number Info:
- Format: Vector of length 6920
- Content: Detection number linking camera and person ID
- Total Detections: Overall count across all cameras
- Total Persons: Number of unique individuals in the dataset
- Dataset Name: The dataset's official name
Need downstream help?
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.