SimuGAN: Unsupervised forward modeling and optimal design of a LIDAR Camera
This work offers a strong specific gain in depth estimation accuracy and stability for LIDAR camera designers and users, particularly for short-distance, energy-saving applications.
The paper addresses the issue of inaccurate and unreliable depth estimation in low-power LIDAR cameras due to noisy back-scattered pulses. It proposes an unsupervised forward model using GANs to simulate the camera's behavior and then optimizes camera parameters for improved depth, accuracy, and stability.
Energy-saving LIDAR camera for short distances estimates an object's distance using temporally intensity-coded laser light pulses and calculates the maximum correlation with the back-scattered pulse. Though on low power, the backs-scattered pulse is noisy and unstable, which leads to inaccurate and unreliable depth estimation. To address this problem, we use GANs (Generative Adversarial Networks), which are two neural networks that can learn complicated class distributions through an adversarial process. We learn the LIDAR camera's hidden properties and behavior, creating a novel, fully unsupervised forward model that simulates the camera. Then, we use the model's differentiability to explore the camera parameter space and optimize those parameters in terms of depth, accuracy, and stability. To achieve this goal, we also propose a new custom loss function designated to the back-scattered code distribution's weaknesses and its circular behavior. The results are demonstrated on both synthetic and real data.