UWaterloo Precise Synthetic Image and LiDAR (PreSIL) Dataset

The PreSIL dataset consists of over 50,000 instances and includes high-definition images with full resolution depth information, semantic segmentation (images), point-wise segmentation (point clouds), ground point labels (point clouds), and detailed annotations for all vehicles and people.

Data and Resources

Additional Info

Field Value
Source Waterloo University - Intelligent Systems Engineering Lab
Maintainer UWaterloo Intelligent Systems Engineering Lab
Associated Tasks Classification and Detection; Object Instance Segmentation; Semantic Segmentation
Geographical Area Waterloo, Canada
Number of Instances 50,000+ instances
Package Description The package provides a large (50,000+ frames) dataset in the KITTI format with extended information and labels. For each data frame, the dataset contains: – A 1920 x 1080 resolution color image. – A 1920 x 1080 resolution depth map. – For each object in the image: ∗ A label in the KITTI 3D Object Detection format (2D and 3D bounding boxes, occlusion, truncation, and class information). ∗ An augmented label with entity ID, the number of 2D points in the image associated with the object, speed, pitch, roll, and a model name. – A 1920 x 1080 instance segmentation image for vehicles and people (corresponding to entity ID in label), semantic segmentation for all other pixels. – An instantaneous point cloud without any motion distortion in a forward-facing 90 degree FoV. Similar to the KITTI format but without reflectance values. For each point there is also an associated entity ID if the point corresponds to a vehicle or person. – Intrinsic and extrinsic calibration information for all sensors.
Dataset has missing values False