JIST: Joint Image and Sequence Training for Sequential Visual Place Recognition.
Gabriele Moreno BertonGabriele TrivignoBarbara CaputoCarlo MasonePublished in: CoRR (2024)
Keyphrases
- single image
- image classification
- image content
- visual perception
- multiscale
- image features
- low level
- image data
- template matching
- visual cues
- input image
- image retrieval
- human visual
- visual appearance
- image segmentation
- visual data
- image representation
- spatial relations
- image pixels
- spatial information
- spatial layout
- image analysis
- visually similar
- visual properties
- keypoints
- image collections
- visual information
- pixel values
- neural network
- visual features
- feature points
- visual similarity
- high resolution
- similarity measure
- image sequences
- image processing