Video Frame Interpolation

Video frame interpolation (VFI) aims to generate realistic intermediate frames between existing ones in a video sequence, increasing frame rate and improving visual smoothness. Current research heavily focuses on improving accuracy and efficiency, exploring various model architectures including convolutional neural networks, transformers, and diffusion models, often incorporating techniques like optical flow estimation, motion modeling, and multi-modal data fusion (e.g., combining RGB and event camera data). These advancements have significant implications for applications such as slow-motion video generation, video upscaling, and enhancing the quality of existing video content, driving improvements in both entertainment and scientific visualization.