Deep Active Learning with Noisy Oracle in Object Detection.
Marius SchubertTobias RiedlingerKarsten KahlMatthias RottmannPublished in: VISIGRAPP (2): VISAPP (2024)
Keyphrases
- object detection
- active learning
- label noise
- face detection
- learning algorithm
- computer vision
- object class
- object recognition
- pedestrian detection
- object classification
- selective sampling
- machine learning
- object categories
- multi class
- training set
- human detection
- pool based active learning
- random sampling
- imbalanced data classification
- oracle database
- noisy data
- transfer learning
- missing data
- training examples
- background subtraction
- learning strategies
- deep learning
- crowd sourced
- active learning strategies
- experimental design
- sample complexity
- scene understanding
- noisy environments
- supervised learning
- database