How Well Do Vision Transformers (VTs) Transfer to the Non-natural Image Domain? An Empirical Study Involving Art Classification.
Vincent TonkesMatthia SabatelliPublished in: ECCV Workshops (1) (2022)
Keyphrases
- natural images
- multiscale
- denoising
- image classification
- computer vision
- feature extraction
- classification accuracy
- machine learning
- decision trees
- cross domain
- support vector
- training set
- support vector machine
- ground truth
- image processing
- transfer learning
- image denoising
- natural scenes
- feature space
- feature vectors
- text classification
- feature selection
- test images