Close
Finding Your (3D) Center: 3D object detection using a learned loss
To develop a new optimization procedure that allows training for 3D detection with raw 3D scans while using as little as 5 % of the object labels and still achieving comparable performance.
22 July 2020
Abstract
We present a novel approach to training for point cloud object detection. By employing two networks, we show how first a smaller local network can be trained with a fraction of the required training data. We then demonstrate how this first network can be used as a loss function to train a second full-scene network without the need for labels.
Publication: PDF