Better video object segmentation with few examples by using consistency over time with TTI
Better video object segmentation with few examples by using consistency over time with TTI
Temporal Transductive Inference for Few-Shot Video Object Segmentation
arXiv paper abstract https://arxiv.org/abs/2203.14308v1
arXiv PDF paper https://arxiv.org/pdf/2203.14308v1.pdf
GitHub https://github.com/MSiam/tti_fsvos
Few-shot video object segmentation (FS-VOS) aims at segmenting video frames using a few labelled examples of classes not seen during initial training.
… present … temporal transductive inference (TTI) approach that leverages temporal consistency in the unlabelled video frames during few-shot inference.
… approach is the use of both global and local temporal constraints.
The objective of the global constraint is to learn consistent linear classifiers for novel classes across the image sequence, … … the local constraint enforces the proportion of foreground/background regions in each frame to be coherent across a local temporal window.
These constraints act as spatiotemporal regularizers during the transductive inference to increase temporal coherence and reduce overfitting on the few-shot support set.
… model outperforms state-of-the-art meta-learning approaches in terms of mean intersection over union on YouTube-VIS by 2.8%. …
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
LinkedIn https://www.linkedin.com/in/morris-lee-47877b7b
