CU-Net: Coupled U-Nets
This work addresses parameter efficiency in neural networks for computer vision tasks like human pose estimation, but it appears incremental as it builds on existing U-Net architectures.
The paper tackles the problem of parameter efficiency in U-Net architectures for human pose estimation by introducing CU-Net, which couples stacked U-Nets through connections between their semantic blocks. The result is a model that achieves comparable accuracy to state-of-the-art methods while using at least 60% fewer parameters.
We design a new connectivity pattern for the U-Net architecture. Given several stacked U-Nets, we couple each U-Net pair through the connections of their semantic blocks, resulting in the coupled U-Nets (CU-Net). The coupling connections could make the information flow more efficiently across U-Nets. The feature reuse across U-Nets makes each U-Net very parameter efficient. We evaluate the coupled U-Nets on two benchmark datasets of human pose estimation. Both the accuracy and model parameter number are compared. The CU-Net obtains comparable accuracy as state-of-the-art methods. However, it only has at least 60% fewer parameters than other approaches.