Greit-HRNet: Grouped Lightweight High-Resolution Network for Human Pose Estimation
This work addresses efficiency and accuracy challenges in human pose estimation for computer vision applications, representing an incremental improvement over existing lightweight methods.
The paper tackled the problem of maintaining weight consistency and capturing global spatial information in lightweight high-resolution networks for human pose estimation by proposing Greit-HRNet with Grouped Channel Weighting and Global Spatial Weighting blocks, achieving superior performance on MS-COCO and MPII datasets compared to other state-of-the-art lightweight networks.
As multi-scale features are necessary for human pose estimation tasks, high-resolution networks are widely applied. To improve efficiency, lightweight modules are proposed to replace costly point-wise convolutions in high-resolution networks, including channel weighting and spatial weighting methods. However, they fail to maintain the consistency of weights and capture global spatial information. To address these problems, we present a Grouped lightweight High-Resolution Network (Greit-HRNet), in which we propose a Greit block including a group method Grouped Channel Weighting (GCW) and a spatial weighting method Global Spatial Weighting (GSW). GCW modules group conditional channel weighting to make weights stable and maintain the high-resolution features with the deepening of the network, while GSW modules effectively extract global spatial information and exchange information across channels. In addition, we apply the Large Kernel Attention (LKA) method to improve the whole efficiency of our Greit-HRNet. Our experiments on both MS-COCO and MPII human pose estimation datasets demonstrate the superior performance of our Greit-HRNet, outperforming other state-of-the-art lightweight networks.