Efficient Deep Demosaicing with Spatially Downsampled Isotropic Networks
This work addresses the need for lightweight demosaicing methods for mobile imaging applications, but it is incremental as it builds on existing isotropic network designs.
The paper tackles the problem of computationally expensive deep learning for image demosaicing on mobile platforms by proposing isotropic networks with spatial downsampling, resulting in improved efficiency and strong performance on demosaicing and joint-demosaicing-and-denoising tasks.
In digital imaging, image demosaicing is a crucial first step which recovers the RGB information from a color filter array (CFA). Oftentimes, deep learning is utilized to perform image demosaicing. Given that most modern digital imaging applications occur on mobile platforms, applying deep learning to demosaicing requires lightweight and efficient networks. Isotropic networks, also known as residual-in-residual networks, have been often employed for image demosaicing and joint-demosaicing-and-denoising (JDD). Most demosaicing isotropic networks avoid spatial downsampling entirely, and thus are often prohibitively expensive computationally for mobile applications. Contrary to previous isotropic network designs, this paper claims that spatial downsampling to a signficant degree can improve the efficiency and performance of isotropic networks. To validate this claim, we design simple fully convolutional networks with and without downsampling using a mathematical architecture design technique adapted from DeepMAD, and find that downsampling improves empirical performance. Additionally, empirical testing of the downsampled variant, JD3Net, of our fully convolutional networks reveals strong empirical performance on a variety of image demosaicing and JDD tasks.