Cutting-Edge Techniques for Depth Map Super-Resolution
This addresses a practical computer vision problem for applications using depth sensors, but it is incremental as it builds on existing joint image filtering and CNN approaches.
The paper tackles the problem of low-resolution depth maps from sensors by proposing a novel joint image filtering algorithm using a Swin transformer and a Nonlinear Activation Free network, achieving state-of-the-art performance with competitive computation time for noisy depth map super-resolution.
To overcome hardware limitations in commercially available depth sensors which result in low-resolution depth maps, depth map super-resolution (DMSR) is a practical and valuable computer vision task. DMSR requires upscaling a low-resolution (LR) depth map into a high-resolution (HR) space. Joint image filtering for DMSR has been applied using spatially-invariant and spatially-variant convolutional neural network (CNN) approaches. In this project, we propose a novel joint image filtering DMSR algorithm using a Swin transformer architecture. Furthermore, we introduce a Nonlinear Activation Free (NAF) network based on a conventional CNN model used in cutting-edge image restoration applications and compare the performance of the techniques. The proposed algorithms are validated through numerical studies and visual examples demonstrating improvements to state-of-the-art performance while maintaining competitive computation time for noisy depth map super-resolution.