GAN-Based LiDAR Intensity Simulation
This work addresses the need for efficient sensor simulation in autonomous driving development, though it appears incremental by building on existing GAN-based translation approaches.
The paper tackled the problem of realistic LiDAR sensor simulation for autonomous driving by developing a GAN-based method that translates camera images to LiDAR point clouds, using segmentation and depth maps as additional inputs and evaluating realism through object detection generalization, demonstrating successful simulation.
Realistic vehicle sensor simulation is an important element in developing autonomous driving. As physics-based implementations of visual sensors like LiDAR are complex in practice, data-based approaches promise solutions. Using pairs of camera images and LiDAR scans from real test drives, GANs can be trained to translate between them. For this process, we contribute two additions. First, we exploit the camera images, acquiring segmentation data and dense depth maps as additional input for training. Second, we test the performance of the LiDAR simulation by testing how well an object detection network generalizes between real and synthetic point clouds to enable evaluation without ground truth point clouds. Combining both, we simulate LiDAR point clouds and demonstrate their realism.