R-FCN Object Detection Ensemble based on Object Resolution and Image Quality.
Christoffer Bøgelund RasmussenKamal NasrollahiThomas B. MoeslundPublished in: IJCCI (2017)
Keyphrases
- image quality
- object detection
- image resolution
- visual quality
- object segmentation
- d objects
- bit rate
- category level
- human visual system
- compression ratio
- image data
- object categories
- object recognition
- object detection and pose estimation
- quality assessment
- object detectors
- contextual cues
- face detection
- high quality
- bounding box
- learning algorithm
- computer vision
- object hypotheses
- scene understanding
- object class
- object tracking
- ensemble learning
- reconstructed image
- object model
- high resolution
- video sequences
- error diffusion
- pedestrian detection
- ensemble methods
- object models
- training data
- image quality metrics
- image features
- feature selection
- gray level
- modulation transfer function
- background subtraction