How Well Do Vision Transformers (VTs) Transfer To The Non-Natural Image Domain? An Empirical Study Involving Art Classification.
Vincent TonkesMatthia SabatelliPublished in: CoRR (2022)
Keyphrases
- natural images
- cross domain
- image denoising
- computer vision
- denoising
- natural scenes
- transfer learning
- feature space
- image classification
- support vector machine
- power spectrum
- superpixels
- higher order
- classification accuracy
- training set
- multiscale
- vision system
- augmented reality
- decision trees
- feature selection
- ground truth
- input image
- feature vectors
- object recognition
- support vector
- feature extraction
- visual patterns
- machine learning