Bayesian Triplet Loss: Uncertainty Quantification in Image Retrieval.
Frederik WarburgMartin JørgensenJavier CiveraSøren HaubergPublished in: ICCV (2021)
Keyphrases
- image retrieval
- decision theory
- dempster shafer
- image database
- multiple feature fusion
- relevance feedback
- inherent uncertainty
- data driven
- retrieval accuracy
- nearest neighbor search
- retrieval method
- belief functions
- image matching
- image representation
- visual features
- cross media
- uncertain data
- image search
- bayesian networks
- image content
- maximum likelihood
- neural network
- image annotation
- bayesian decision
- bayesian analysis
- feature space
- retrieval precision
- high dimensional
- uncertain information
- multimedia retrieval
- utility function
- bayesian methods
- text retrieval
- quantitative evaluation
- conditional probabilities
- bayesian inference
- posterior probability
- data sets