RPSNet

Reported on 4 benchmarks across 2 tasks · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Methodology4 results

Incremental LearningonCIFAR-100-B0(5steps of 20 classes)
Average Incremental Accuracy· 2019-06-03
70.5
best: 79.23 (View-Batch(TCIL))
An Adaptive Random Path Selection Approach for Incremental Learning arXiv:1906.01120
Incremental LearningonImageNet100 - 10 steps
Average Incremental Accuracy Top-5· 2019-06-03
87.9
best: 94.17 (TCIL)
An Adaptive Random Path Selection Approach for Incremental Learning arXiv:1906.01120
Incremental LearningonImageNet100 - 10 steps
Final Accuracy Top-5· 2019-06-03
74
best: 88.84 (TCIL)
An Adaptive Random Path Selection Approach for Incremental Learning arXiv:1906.01120
Continual LearningonF-CelebA (10 tasks)
Acc
0.5545
best: 0.7564 (CAT (CNN backbone))