RoBERTa-CoA: RoBERTa-Based Effective Finetuning Method Using Co-Attention.
Jeong-Hoon KimSungwook ParkJun-Yeong KimJun ParkSe-Hoon JungChun-Bo SimPublished in: IEEE Access (2023)
Keyphrases
- computational cost
- high accuracy
- computationally efficient
- significant improvement
- support vector machine
- mutual information
- main contribution
- synthetic data
- image processing
- dynamic programming
- high precision
- segmentation method
- experimental study
- error rate
- convergence rate
- machine learning
- detection algorithm
- segmentation algorithm
- em algorithm
- input image
- video sequences