Uncertainty for Identifying Open-Set Errors in Visual Object Detection.
Dimity MillerNiko SünderhaufMichael MilfordFeras DayoubPublished in: IEEE Robotics Autom. Lett. (2022)
Keyphrases
- object detection
- measurement errors
- category level
- object recognition
- computer vision
- object classification
- visual features
- visual data
- inherent uncertainty
- scene understanding
- object class
- multi class
- low level
- pedestrian detection
- neural network
- face detection
- visual perception
- scene recognition
- measurement error
- contextual cues
- prediction error
- visual cues
- error analysis
- keywords
- high level
- image segmentation