A Ranking-based, Balanced Loss Function Unifying Classification and Localisation in Object Detection.
Kemal OksuzBaris Can CamEmre AkbasSinan KalkanPublished in: NeurIPS (2020)
Keyphrases
- loss function
- learning to rank
- support vector
- hinge loss
- object detection
- classification accuracy
- bayes rule
- pairwise
- feature vectors
- image classification
- logistic regression
- ranking functions
- ranking algorithm
- web search
- model selection
- reproducing kernel hilbert space
- decision trees
- empirical risk
- feature extraction
- ranking svm
- cross validation
- training samples
- support vector machine
- classification algorithm
- pattern classification
- feature space
- density estimation
- generalization ability
- potential functions
- text classification
- supervised learning
- solution path
- active learning