Geology

Point cloud

Pattern recognition (psychology)

Inference

Mathematics

Convolution (computer science)

Economic growth

Lidar

RGB color model

Ranging

Orientation (vector space)

Artificial intelligence

Economics

Geometry

Object detection

Convergence (economics)

Computer science

Computer vision

Artificial neural network

Remote sensing

Telecommunications

Convolutional neural network

Sensors

LiDAR-based or RGB-D-based object detection is used in numerous applications, ranging from autonomous driving to robot vision. Voxel-based 3D convolutional networks have been used for some time to enhance the retention of information when processing point cloud LiDAR data. However, problems remain, including a slow inference speed and low orientation estimation performance. We therefore investigate an improved sparse convolution method for such networks, which significantly increases the speed of both training and inference. We also introduce a new form of angle loss regression to improve the orientation estimation performance and a new data augmentation approach that can enhance the convergence speed and performance. The proposed network produces state-of-the-art results on the KITTI 3D object detection benchmarks while maintaining a fast inference speed.

SECOND: Sparsely Embedded Convolutional Detection