IMDeception: Grouped Information Distilling Super-Resolution Network
This work addresses the practical deployment challenge of SISR models for edge computing by offering a more efficient network structure, though it is incremental as it builds upon existing IMDN mechanisms.
The paper tackles the computational inefficiency of single-image super-resolution (SISR) networks by proposing IMDeception, which uses a Global Progressive Refinement Module and Grouped Information Distilling Blocks to reduce parameters and FLOPS while maintaining performance on par with state-of-the-art models, as demonstrated by real-time deployment on an NVIDIA Jetson Xavier AGX edge device.
Single-Image-Super-Resolution (SISR) is a classical computer vision problem that has benefited from the recent advancements in deep learning methods, especially the advancements of convolutional neural networks (CNN). Although state-of-the-art methods improve the performance of SISR on several datasets, direct application of these networks for practical use is still an issue due to heavy computational load. For this purpose, recently, researchers have focused on more efficient and high-performing network structures. Information multi-distilling network (IMDN) is one of the highly efficient SISR networks with high performance and low computational load. IMDN achieves this efficiency with various mechanisms such as Intermediate Information Collection (IIC), working in a global setting, Progressive Refinement Module (PRM), and Contrast Aware Channel Attention (CCA), employed in a local setting. These mechanisms, however, do not equally contribute to the efficiency and performance of IMDN. In this work, we propose the Global Progressive Refinement Module (GPRM) as a less parameter-demanding alternative to the IIC module for feature aggregation. To further decrease the number of parameters and floating point operations persecond (FLOPS), we also propose Grouped Information Distilling Blocks (GIDB). Using the proposed structures, we design an efficient SISR network called IMDeception. Experiments reveal that the proposed network performs on par with state-of-the-art models despite having a limited number of parameters and FLOPS. Furthermore, using grouped convolutions as a building block of GIDB increases room for further optimization during deployment. To show its potential, the proposed model was deployed on NVIDIA Jetson Xavier AGX and it has been shown that it can run in real-time on this edge device