SOTAVerified

Continual Learning

Continual Learning (also known as Incremental Learning, Life-long Learning) is a concept to learn a model for a large number of tasks sequentially without forgetting knowledge obtained from the preceding tasks, where the data in the old tasks are not available anymore during training new ones.
If not mentioned, the benchmarks here are Task-CL, where task-id is provided on validation.

Source:
Continual Learning by Asymmetric Loss Approximation with Single-Side Overestimation
Three scenarios for continual learning
Lifelong Machine Learning
Continual lifelong learning with neural networks: A review

Papers

Showing 251–300 of 2644 papers

TitleStatusHype
Achieving a Better Stability-Plasticity Trade-off via Auxiliary Networks in Continual LearningCode1
TARGET: Federated Class-Continual Learning via Exemplar-Free DistillationCode1
Preventing Zero-Shot Transfer Degradation in Continual Learning of Vision-Language ModelsCode1
Predictive Experience Replay for Continual Visual Control and ForecastingCode1
SLCA: Slow Learner with Classifier Alignment for Continual Learning on a Pre-trained ModelCode1
Multimodal Parameter-Efficient Few-Shot Class Incremental LearningCode1
GradMA: A Gradient-Memory-based Accelerated Federated Learning with Alleviated Catastrophic ForgettingCode1
A Neural Span-Based Continual Named Entity Recognition ModelCode1
New Insights for the Stability-Plasticity Dilemma in Online Continual LearningCode1
Efficiency 360: Efficient Vision TransformersCode1
Task-Specific Skill Localization in Fine-tuned Language ModelsCode1
Generalized Few-Shot Continual Learning with Contrastive Mixture of AdaptersCode1
Lightweight Transformers for Clinical Natural Language ProcessingCode1
Continual Learning with Scaled Gradient ProjectionCode1
Real-Time Evaluation in Online Continual Learning: A New HopeCode1
Neuro-Symbolic Continual Learning: Knowledge, Reasoning Shortcuts and Concept RehearsalCode1
Open-VCLIP: Transforming CLIP to an Open-vocabulary Video Model via Interpolated Weight OptimizationCode1
IM-IAD: Industrial Image Anomaly Detection Benchmark in ManufacturingCode1
A Comprehensive Survey of Continual Learning: Theory, Method and ApplicationCode1
Progressive Prompts: Continual Learning for Language ModelsCode1
Towards Continual Egocentric Activity Recognition: A Multi-modal Egocentric Activity Dataset for Continual LearningCode1
NAPA-VQ: Neighborhood-Aware Prototype Augmentation with Vector Quantization for Continual LearningCode1
Class-incremental Continual Learning for Instance Segmentation with Image-level Weak SupervisionCode1
CTP:Towards Vision-Language Continual Pretraining via Compatible Momentum Contrast and Topology PreservationCode1
VQACL: A Novel Visual Question Answering Continual Learning SettingCode1
Sparse Coding in a Dual Memory System for Lifelong LearningCode1
Re-basin via implicit Sinkhorn differentiationCode1
3rd Continual Learning Workshop Challenge on Egocentric Category and Instance Level Object UnderstandingCode1
A soft nearest-neighbor framework for continual semi-supervised learningCode1
Lifelong Person Re-Identification via Knowledge Refreshing and ConsolidationCode1
Isolation and Impartial Aggregation: A Paradigm of Incremental Learning without InterferenceCode1
The Effectiveness of World Models for Continual Reinforcement LearningCode1
BeGin: Extensive Benchmark Scenarios and An Easy-to-use Framework for Graph Continual LearningCode1
Wild-Time: A Benchmark of in-the-Wild Distribution Shift over TimeCode1
CODA-Prompt: COntinual Decomposed Attention-based Prompting for Rehearsal-Free Continual LearningCode1
Exemplar-free Continual Learning of Vision Transformers via Gated Class-Attention and Cascaded Feature Drift CompensationCode1
NEVIS'22: A Stream of 100 Tasks Sampled from 30 Years of Computer Vision ResearchCode1
Mining Unseen Classes via Regional Objectness: A Simple Baseline for Incremental SegmentationCode1
A Theoretical Study on Solving Continual LearningCode1
Do Pre-trained Models Benefit Equally in Continual Learning?Code1
Broken Neural Scaling LawsCode1
Attribution-aware Weight Transfer: A Warm-Start Initialization for Class-Incremental Semantic SegmentationCode1
On the Effectiveness of Lipschitz-Driven Rehearsal in Continual LearningCode1
Continual Learning by Modeling Intra-Class VariationCode1
Towards Continual Adaptation in Industrial Anomaly DetectionCode1
Patch-based Knowledge Distillation for Lifelong Person Re-IdentificationCode1
Distilling Causal Effect from Miscellaneous Other-Class for Continual Named Entity RecognitionCode1
CLAD: A realistic Continual Learning benchmark for Autonomous DrivingCode1
CLIP model is an Efficient Continual LearnerCode1
Understanding Collapse in Non-Contrastive Siamese Representation LearningCode1
Show:102550
← PrevPage 6 of 53Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1Multi-task Learning (MTL; Upper Bound)F1 - macro0.88—Unverified
2CTRF1 - macro0.84—Unverified
3B-CLF1 - macro0.81—Unverified
4LAMOLF1 - macro0.81—Unverified
5OWMF1 - macro0.79—Unverified
6A-GEMF1 - macro0.78—Unverified
7HATF1 - macro0.78—Unverified
8Independent Learning (ONE)F1 - macro0.78—Unverified
9KANF1 - macro0.77—Unverified
10Naive Continual Learning (NCL)F1 - macro0.77—Unverified
#ModelMetricClaimedVerifiedStatus
1NetTailordecathlon discipline (Score)3,744—Unverified
2Depthwise Soft Sharingdecathlon discipline (Score)3,507—Unverified
3Parallel Res. adapt.decathlon discipline (Score)3,412—Unverified
4Depthwise Sharingdecathlon discipline (Score)3,234—Unverified
5Series Res. adapt.decathlon discipline (Score)3,159—Unverified
6Res. adapt. (large)decathlon discipline (Score)3,131—Unverified
7DANdecathlon discipline (Score)2,851—Unverified
8Piggybackdecathlon discipline (Score)2,838—Unverified
9Res. adapt. finetune alldecathlon discipline (Score)2,643—Unverified
10Res. adapt. decaydecathlon discipline (Score)2,621—Unverified
#ModelMetricClaimedVerifiedStatus
1Model Zoo-ContinualAverage Accuracy94.99—Unverified
2ALTAAverage Accuracy92.98—Unverified
3kNN-CLIPAverage Accuracy90.8—Unverified
4RMNAverage Accuracy81—Unverified
5CPGAverage Accuracy80.9—Unverified
6CondConvContinualAverage Accuracy77.4—Unverified
7PAENetAverage Accuracy77.1—Unverified
8CPG-lightAverage Accuracy77—Unverified
9PackNetAverage Accuracy67.5—Unverified
#ModelMetricClaimedVerifiedStatus
1ALTA-ViTB/16Average Accuracy89.8—Unverified
2ALTA-RN50x4Average Accuracy84.73—Unverified
3ALTA-RN101Average Accuracy83.35—Unverified
4ALTA-RN50Average Accuracy81.07—Unverified
5SNCLAverage Accuracy52.85—Unverified
6DER [buzzega2020dark]Average Accuracy51.78—Unverified
7ER[riemer2018learning]Average Accuracy48.64—Unverified
8iCaRL [rebuffi2017icarl]Average Accuracy31.55—Unverified
9A-GEM [chaudhry2018efficient]Average Accuracy25.33—Unverified
#ModelMetricClaimedVerifiedStatus
1CAT (CNN backbone)Acc0.76—Unverified
2CAT (MLP backbone)Acc0.69—Unverified
3EWCAcc0.65—Unverified
4HyperNetAcc0.6—Unverified
5PathNetAcc0.58—Unverified
6HATAcc0.57—Unverified
7RPSNetAcc0.55—Unverified
#ModelMetricClaimedVerifiedStatus
1CTRF1 - macro0.95—Unverified
2HATF1 - macro0.95—Unverified
3CATF1 - macro0.95—Unverified
4B-CLF1 - macro0.95—Unverified
5EWCF1 - macro0.92—Unverified
6LAMOLF1 - macro0.46—Unverified
#ModelMetricClaimedVerifiedStatus
1CondConvContinualAccuracy84.26—Unverified
2H$^{2}$Accuracy84.1—Unverified
3CPGAccuracy83.59—Unverified
4PiggybackAccuracy80.5—Unverified
5PackNetAccuracy80.41—Unverified
6ProgressiveNetAccuracy78.94—Unverified
#ModelMetricClaimedVerifiedStatus
1CTRF1 - macro0.89—Unverified
2CATF1 - macro0.87—Unverified
3HATF1 - macro0.86—Unverified
4KANF1 - macro0.81—Unverified
5B-CLF1 - macro0.77—Unverified
6EWCF1 - macro0.66—Unverified
#ModelMetricClaimedVerifiedStatus
1CondConvContinualAccuracy97.16—Unverified
2CPGAccuracy96.62—Unverified
3H$^{2}$Accuracy94.9—Unverified
4PiggybackAccuracy94.77—Unverified
5ProgressiveNetAccuracy93.41—Unverified
6PackNetAccuracy93.04—Unverified
#ModelMetricClaimedVerifiedStatus
1PiggybackAccuracy76.16—Unverified
2ProgressiveNetAccuracy76.16—Unverified
3CondConvContinualAccuracy76.16—Unverified
4CPGAccuracy75.81—Unverified
5PackNetAccuracy75.71—Unverified
6H$^{2}$Accuracy75.71—Unverified
#ModelMetricClaimedVerifiedStatus
1CondConvContinualAccuracy80.77—Unverified
2CPGAccuracy80.33—Unverified
3PiggybackAccuracy79.91—Unverified
4ProgressiveNetAccuracy76.35—Unverified
5H$^{2}$Accuracy76.2—Unverified
6PackNetAccuracy76.17—Unverified
#ModelMetricClaimedVerifiedStatus
1CPGAccuracy92.8—Unverified
2CondConvContinualAccuracy92.61—Unverified
3H$^{2}$Accuracy90.6—Unverified
4PiggybackAccuracy89.62—Unverified
5ProgressiveNetAccuracy89.21—Unverified
6PackNetAccuracy86.11—Unverified
#ModelMetricClaimedVerifiedStatus
1CondConvContinualAccuracy78.32—Unverified
2CPGAccuracy77.15—Unverified
3H$^{2}$Accuracy75.1—Unverified
4ProgressiveNetAccuracy74.94—Unverified
5PiggybackAccuracy71.33—Unverified
6PackNetAccuracy69.4—Unverified
#ModelMetricClaimedVerifiedStatus
1ALTA-ViTB/16Average Accuracy92.85—Unverified
2ALTA-RN50x4Average Accuracy84.91—Unverified
3RMN (Resnet)Average Accuracy84.9—Unverified
4ALTA-RN101Average Accuracy84.77—Unverified
5ALTA-RN50Average Accuracy83.87—Unverified
#ModelMetricClaimedVerifiedStatus
1RMNAccuracy68.1—Unverified
2CondConvContinualAccuracy61.32—Unverified
3CCGNAccuracy35.24—Unverified
4DGMwAccuracy17.82—Unverified
5DGMaAccuracy15.16—Unverified
#ModelMetricClaimedVerifiedStatus
1RMNAverage Accuracy97.99—Unverified
2Model Zoo-ContinualAverage Accuracy97.71—Unverified
3CODE-CLAverage Accuracy96.56—Unverified
#ModelMetricClaimedVerifiedStatus
1TAG-RMSPropAccuracy62.59—Unverified
#ModelMetricClaimedVerifiedStatus
1CODE-CLAverage Accuracy93.32—Unverified
#ModelMetricClaimedVerifiedStatus
1TEST1:3 Accuracy2—Unverified
#ModelMetricClaimedVerifiedStatus
1TAG-RMSPropAverage Accuracy62.79—Unverified
#ModelMetricClaimedVerifiedStatus
1IBMAccuracy82.69—Unverified
#ModelMetricClaimedVerifiedStatus
1IBMAccuracy88.15—Unverified
#ModelMetricClaimedVerifiedStatus
1Model Zoo-ContinualAverage Accuracy84.27—Unverified
#ModelMetricClaimedVerifiedStatus
1TAG-RMSPropAccuracy61.58—Unverified
#ModelMetricClaimedVerifiedStatus
1CODE-CLAverage Accuracy68.83—Unverified
#ModelMetricClaimedVerifiedStatus
1TAG-RMSPropAccuracy57.2—Unverified
#ModelMetricClaimedVerifiedStatus
1IBMAccuracy53.9—Unverified
#ModelMetricClaimedVerifiedStatus
1MRMAcc78.4—Unverified
#ModelMetricClaimedVerifiedStatus
1Model Zoo-ContinualAverage Accuracy99.66—Unverified
#ModelMetricClaimedVerifiedStatus
1CODE-CLAverage Accuracy77.21—Unverified
#ModelMetricClaimedVerifiedStatus
1H$^{2}$Top 1 Accuracy %97.3—Unverified
#ModelMetricClaimedVerifiedStatus
1H$^{2}$Top 1 Accuracy %99.9—Unverified
#ModelMetricClaimedVerifiedStatus
1IBMAccuracy52.38—Unverified