Explain and improve: LRP-inference fine-tuning for image captioning models.
Jiamei SunSebastian LapuschkinWojciech SamekAlexander BinderPublished in: Inf. Fusion (2022)
Keyphrases
- fine tuning
- random fields
- image data
- image analysis
- input image
- image classification
- fine tuned
- image collections
- image content
- single image
- parametric models
- edge detection
- multiscale
- bayesian framework
- template matching
- high resolution
- segmentation method
- image features
- image representation
- image segmentation
- test images
- structured prediction
- image regions
- region of interest
- image pixels
- viable alternative
- contrast enhancement
- keypoints
- image set
- parameter estimation
- visual features
- feature points
- similarity measure