Copyright (C) 2023 Jing Wang
The GRC optimization problem
Four experiments were conducted to compare GRC with LRC, CRC, and SRC on four benchmark face databases including the AR, FEI, FERET, and UMIST face databases.
Experiment 1: Tune the parameters
Experiment 2: Compare GRC with LRC, CRC, and SRC when principal components with different percentages (90%, 95%, 98%) of explained variance are extracted.
Experiment 3: Tune the parameters
Experiment 4: Compare GRC with LRC, CRC, and SRC when different numbers of principal components (10:10:300) are extracted.
- Run load_data.m to split the face databases into training and testing sets.
- Run the following scripts sequentially:
- exp_1_classify.m
- exp_1_analysis.m
- exp_2_classify.m
- exp_2_analysis.m
- exp_3_classify.m
- exp_3_analysis.m
- exp_4_classify.m
- exp_4_analysis.m
- More benchmark face databases are refered to: https://github.com/yuzhounh/Face_databases.
Jing Wang
[email protected]
2023-11-23 10:19:28