HyperAI초신경

홈 뉴스 연구 논문 튜토리얼 데이터셋 백과사전 SOTA LLM 모델 GPU 랭킹 컨퍼런스

한국어

HyperAI초신경

Incremental Learning On Cifar 100 50 Classes 2

평가 지표

Average Incremental Accuracy

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	Average Incremental Accuracy	Paper Title	Repository
FOSTER	67.95	FOSTER: Feature Boosting and Compression for Class-Incremental Learning
BiC	53.21	Large Scale Incremental Learning
TCIL-Lite	73.50	Resolving Task Confusion in Dynamic Expansion Architectures for Class Incremental Learning
RMM (Modified ResNet-32)	67.61	RMM: Reinforced Memory Management for Class-Incremental Learning
PODNet (CNN)	63.19	PODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning
D3Former	70.94	D3Former: Debiased Dual Distilled Transformer for Incremental Learning
DER(Standard ResNet-18)	72.45	DER: Dynamically Expandable Representation for Class Incremental Learning
iCaRL*	52.57	iCaRL: Incremental Classifier and Representation Learning
TCIL	73.72	Resolving Task Confusion in Dynamic Expansion Architectures for Class Incremental Learning
UCIR (CNN)*	60.18	Learning a Unified Classifier Incrementally via Rebalancing	-
UCIR (NME)*	60.12	Learning a Unified Classifier Incrementally via Rebalancing	-
CCIL-SD	65.86	Essentials for Class Incremental Learning
DER(Modified ResNet-32）	66.36	DER: Dynamically Expandable Representation for Class Incremental Learning

0 of 13 row(s) selected.