Improving Accuracy of Respiratory Rate Estimation by Restoring High Resolution Features With Transformers and Recursive Convolutional Models.
Alicja KwasniewskaMaciej SzankinJacek RuminskiAnthony SarahDavid GambaPublished in: CVPR Workshops (2021)
Keyphrases
- high resolution
- predictive power
- classification accuracy
- prior knowledge
- feature vectors
- parametric models
- parameter estimation
- computational cost
- abstraction levels
- predictive features
- feature extraction
- parameter estimates
- classification models
- training set
- false positives
- error rate
- learning models
- roc curve
- low resolution
- high accuracy
- learned models
- co occurrence
- image features