A Novel Loss Calibration Strategy for Object Detection Networks Training on Sparsely Annotated Pathological Datasets.
Hansheng LiXin HanYuxin KangXiaoshuang ShiMengdi YanZixu TongQirong BuLei CuiJun FengLin YangPublished in: MICCAI (5) (2020)
Keyphrases
- object detection
- pascal voc
- ground truth labels
- discriminatively trained
- training dataset
- computer vision
- echo state networks
- object class
- object categories
- boosted classifiers
- object detectors
- scene understanding
- camera calibration
- training set
- training examples
- social networks
- active learning
- restricted boltzmann machine
- linear svm
- supervised learning
- object classification
- data sets
- face detection
- multi class
- training process
- pedestrian detection
- training and testing data
- sparse representation
- microarray
- test set
- network structure
- video sequences