DS-NeRV: Implicit Neural Video Representation with Decomposed Static and Dynamic Codes
This work addresses video compression and representation for applications like streaming or storage, offering an incremental improvement over prior NeRV methods.
The paper tackles the problem of implicit neural video representation by proposing DS-NeRV, which decomposes videos into static and dynamic codes to improve compression and detail modeling, achieving 31.2 PSNR with 0.35M parameters and outperforming existing methods in downstream tasks.
Implicit neural representations for video (NeRV) have recently become a novel way for high-quality video representation. However, existing works employ a single network to represent the entire video, which implicitly confuse static and dynamic information. This leads to an inability to effectively compress the redundant static information and lack the explicitly modeling of global temporal-coherent dynamic details. To solve above problems, we propose DS-NeRV, which decomposes videos into sparse learnable static codes and dynamic codes without the need for explicit optical flow or residual supervision. By setting different sampling rates for two codes and applying weighted sum and interpolation sampling methods, DS-NeRV efficiently utilizes redundant static information while maintaining high-frequency details. Additionally, we design a cross-channel attention-based (CCA) fusion module to efficiently fuse these two codes for frame decoding. Our approach achieves a high quality reconstruction of 31.2 PSNR with only 0.35M parameters thanks to separate static and dynamic codes representation and outperforms existing NeRV methods in many downstream tasks. Our project website is at https://haoyan14.github.io/DS-NeRV.