PS-NeRV: Patch-wise Stylized Neural Representations for Videos
This work addresses video representation for applications such as compression and inpainting, but it is incremental as it builds on existing INR methods with patch-wise modifications.
The authors tackled the problem of representing videos using implicit neural representations (INRs) by proposing PS-NeRV, a patch-wise method that improves reconstruction performance and decoding speed, achieving excellent results in tasks like video compression and inpainting.
We study how to represent a video with implicit neural representations (INRs). Classical INRs methods generally utilize MLPs to map input coordinates to output pixels. While some recent works have tried to directly reconstruct the whole image with CNNs. However, we argue that both the above pixel-wise and image-wise strategies are not favorable to video data. Instead, we propose a patch-wise solution, PS-NeRV, which represents videos as a function of patches and the corresponding patch coordinate. It naturally inherits the advantages of image-wise methods, and achieves excellent reconstruction performance with fast decoding speed. The whole method includes conventional modules, like positional embedding, MLPs and CNNs, while also introduces AdaIN to enhance intermediate features. These simple yet essential changes could help the network easily fit high-frequency details. Extensive experiments have demonstrated its effectiveness in several video-related tasks, such as video compression and video inpainting.