EgoCart: 9 Egocentric Retail Videos with 3-DOF Pose and Zone Labels

9 MP4 video sequences of egocentric retail store footage captured by shopping cart cameras, created by Voxel51. Each frame includes 3-DOF camera pose (position and orientation) and store-zone classification labels for indoor localization tasks. The data includes synchronized RGB frames and depth maps from a real supermarket environment.

Use Cases

• Indoor localization using the 3-DOF camera pose to track cart movement
• Store zone classification using the store-zone class labels to identify supermarket sections
• Visual odometry research utilizing sequential MP4 video frames and depth maps

Strengths

• 9 distinct recording sequences as MP4 videos
• Frame-level 3-DOF pose annotations (position and orientation)
• Dual-modality data including both RGB and depth maps

Limitations

• Small sample size of 9 video sequences
• Environmental bias to a single supermarket layout
• Perspective limited to shopping cart mounting height

Provenance

Source

Voxel51

Collection Method

Cameras and depth sensors mounted on shopping carts in a real supermarket.

Freshness

Last updated March 2026.

Geography

United States