Adaptive Illumination based Depth Sensing using Deep Superpixel and Soft Sampling Approximation
This work addresses the problem of improving depth sensing accuracy for applications like robotics or autonomous vehicles, but it is incremental as it builds on existing fusion techniques with adaptive sampling.
The paper tackles the challenge of estimating dense depth maps from sparse depth sampling by jointly training an adaptive sparse depth sampling network with an RGB-sparse depth fusion network to generate optimal adaptive sampling masks. The result is that these masks generalize well to various fusion algorithms at low sampling rates, such as 0.0625%.
Dense depth map capture is challenging in existing active sparse illumination based depth acquisition techniques, such as LiDAR. Various techniques have been proposed to estimate a dense depth map based on fusion of the sparse depth map measurement with the RGB image. Recent advances in hardware enable adaptive depth measurements resulting in further improvement of the dense depth map estimation. In this paper, we study the topic of estimating dense depth from depth sampling. The adaptive sparse depth sampling network is jointly trained with a fusion network of an RGB image and sparse depth, to generate optimal adaptive sampling masks. We show that such adaptive sampling masks can generalize well to many RGB and sparse depth fusion algorithms under a variety of sampling rates (as low as $0.0625\%$). The proposed adaptive sampling method is fully differentiable and flexible to be trained end-to-end with upstream perception algorithms.