Class Incremental Learning on cifar100

Metric: 10-stage average accuracy (higher is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Sort:

#	Model↕	10-stage average accuracy▼	Augmentations	Paper	Date↕	Code
1	S&B	68.18	No	Split-and-Bridge: Adaptable Class Incremental Le...	2021-07-03	Code
2	SCR	65.98	No	Supervised Contrastive Learning	2020-04-23	Code
3	iCaRL	63.24	No	iCaRL: Incremental Classifier and Representation...	2016-11-23	Code
4	LUCIR	56.53	No	-	-	Code
5	ABD	54.44	No	Always Be Dreaming: A New Approach for Data-Free...	2021-06-17	Code
6	EMR	48.66	No	On Tiny Episodic Memories in Continual Learning	2019-02-27	Code
7	A-GEM	45.76	No	Efficient Lifelong Learning with A-GEM	2018-12-02	Code

#1S&BSOTA
68.18
10-stage average accuracy· 2021-07-03
Split-and-Bridge: Adaptable Class Incremental Learning within a Single Neural Network Code
#2SCRSOTA
65.98
10-stage average accuracy· 2020-04-23
Supervised Contrastive Learning Code
#3iCaRLSOTA
63.24
10-stage average accuracy· 2016-11-23
iCaRL: Incremental Classifier and Representation Learning Code
#4LUCIR
56.53
10-stage average accuracy
No paperCode
#5ABD
54.44
10-stage average accuracy· 2021-06-17
Always Be Dreaming: A New Approach for Data-Free Class-Incremental Learning Code
#6EMR
48.66
10-stage average accuracy· 2019-02-27
On Tiny Episodic Memories in Continual Learning Code
#7A-GEM
45.76
10-stage average accuracy· 2018-12-02
Efficient Lifelong Learning with A-GEM Code