Video Instance Segmentation with Recurrent Graph Neural Networks.
Joakim JohnanderEmil BrissmanMartin DanelljanMichael FelsbergPublished in: GCPR (2021)
Keyphrases
- neural network
- video segmentation
- feed forward
- recurrent neural networks
- image segmentation
- video data
- video sequences
- medical images
- video streams
- segmentation method
- segmentation algorithm
- video content
- pattern recognition
- level set
- multiscale
- text detection
- multimedia
- object segmentation and tracking
- min cut max flow
- fuzzy logic
- video frames
- connected components
- temporal coherence
- graph representation
- video surveillance
- graph structure
- video analysis
- temporal segmentation
- foreground background segmentation
- normalized cut
- genetic algorithm
- energy function
- structured data
- region growing
- back propagation
- segmentation accuracy
- fully unsupervised
- directed graph
- video objects
- real time
- graph theory
- artificial neural networks
- random walk
- pixel level
- graph model
- low grade gliomas
- topic segmentation
- social networks
- video scene
- region merging
- dynamic textures
- motion segmentation
- self organizing maps