Transforming Limitations into Advantages: Improving Small Object Detection Accuracy with SC-AttentionIoU Loss Function.
Mingle ZhouChangle YiMin LiHonglin WanGang LiDelong HanPublished in: ICANN (7) (2023)
Keyphrases
- loss function
- detection accuracy
- detection algorithm
- face detection
- pairwise
- support vector
- processing speed
- detection rate
- empirical risk
- learning to rank
- number of false alarms
- logistic regression
- regularization term
- false detections
- convex loss functions
- risk minimization
- hinge loss
- reproducing kernel hilbert space
- stochastic gradient descent
- feature extraction
- pairwise constraints
- boosting framework
- true positive rate
- partial occlusion
- boosting algorithms
- machine learning
- recognition rate
- supervised learning
- computational cost
- moving objects
- feature selection