FuSta

Hybrid Neural Fusion for Full-frame Video Stabilization


Overview Video

Abstract
Existing video stabilization methods either require aggressive cropping of frame boundaries or generate distortion artifacts on the stabilized frames. In this work, we present an algorithm for full-frame video stabilization by first estimating dense warp fields. Full-frame stabilized frames can then be synthesized by fusing warped contents from neighboring frames. The core technical novelty lies in our learning-based hybrid-space fusion that alleviates artifacts caused by optical flow inaccuracy and fast-moving objects. We validate the effectiveness of our method on the NUS and selfie video datasets. Extensive experiment results demonstrate the merits of our approach over prior video stabilization methods.
Papers

Citation

Yu-Lun Liu, Wei-Sheng Lai, Ming-Hsuan Yang, Yung-Yu Chuang, and Jia-Bin Huang, "Neural Re-rendering for Full-frame Video Stabilization", arXiv, 2020


Bibtex
@inproceedings{Liu-FuSta-2021,
    author    = {Liu, Yu-Lun and Lai, Wei-Sheng and Yang, Ming-Hsuan and Chuang, Yung-Yu and Huang, Jia-Bin}, 
    title     = {Hybrid Neural Fusion for Full-frame Video Stabilization}, 
    journal   = {arXiv preprint},
    year      = {2021}
}
Download
Results

References