UVMO: Deep unsupervised visual reconstruction-based multimodal-assisted odometry.
Songrui HanMingchi LiHongying TangYaozhe SongGuanjun TongPublished in: Pattern Recognit. (2024)
Keyphrases
- cross modal
- unsupervised learning
- multimodal information
- deep architectures
- visual perception
- visual features
- multi modal
- machine learning
- inertial sensors
- visual information
- data driven
- semi supervised
- supervised learning
- neural network
- three dimensional
- compressive sensing
- visual cues
- reconstruction error
- error analysis
- low level
- weakly supervised
- compressed sensing
- reconstruction process
- high resolution
- multimodal interaction
- audio visual
- deep belief networks
- multimedia
- image reconstruction