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 17011750 of 2644 papers

TitleStatusHype
Towards General Purpose Medical AI: Continual Learning Medical Foundation Model0
Lifelong Machine Learning Potentials0
Dynamic Y-KD: A Hybrid Approach to Continual Instance Segmentation0
On Momentum-Based Gradient Methods for Bilevel Optimization with Nonconvex Lower-Level0
Centroid Distance Distillation for Effective Rehearsal in Continual LearningCode0
Semiparametric Language Models Are Scalable Continual Learners0
Improving Expert Specialization in Mixture of Experts0
Improving Representational Continuity via Continued PretrainingCode0
Better Generative Replay for Continual Federated Learning0
Detachedly Learn a Classifier for Class-Incremental Learning0
Preventing Catastrophic Forgetting in Continual Learning of New Natural Language Tasks0
Effects of Architectures on Continual Semantic Segmentation0
Emphasizing Unseen Words: New Vocabulary Acquisition for End-to-End Speech Recognition0
New Insights on Relieving Task-Recency Bias for Online Class Incremental LearningCode0
À-la-carte Prompt Tuning (APT): Combining Distinct Data Via Composable Prompting0
Error Sensitivity Modulation based Experience Replay: Mitigating Abrupt Representation Drift in Continual LearningCode0
Task-Aware Information Routing from Common Representation Space in Lifelong LearningCode0
Theory on Forgetting and Generalization of Continual Learning0
On the Limitation and Experience Replay for GNNs in Continual Learning0
Utility-based Perturbed Gradient Descent: An Optimizer for Continual Learning0
Efficient Parametric Approximations of Neural Network Function Space Distance0
Stochastic Gradient Descent-Induced Drift of Representation in a Two-Layer Neural Network0
Multi-scale Feature Alignment for Continual Learning of Unlabeled Domains0
Avalanche: A PyTorch Library for Deep Continual Learning0
Online Continual Learning via the Knowledge Invariant and Spread-out Properties0
Continual Learning for Predictive Maintenance: Overview and Challenges0
Restricted Orthogonal Gradient Projection for Continual Learning0
Adversarial Learning Networks: Source-free Unsupervised Domain Incremental Learning0
CyclicFL: A Cyclic Model Pre-Training Approach to Efficient Federated Learning0
Continual Graph Learning: A Survey0
Streaming LifeLong Learning With Any-Time Inference0
Class-Incremental Learning with RepetitionCode0
Probabilistic Bilevel Coreset Selection0
Adaptively Integrated Knowledge Distillation and Prediction Uncertainty for Continual Learning0
Hierarchical Memory Pool Based Edge Semi-Supervised Continual Learning Method0
Knowledge Acquisition and Completion for Long-Term Human-Robot Interactions using Knowledge Graph Embedding0
Artificial Neuronal Ensembles with Learned Context Dependent GatingCode0
Self-recovery of memory via generative replay0
Predictive World Models from Real-World Partial ObservationsCode0
A Distinct Unsupervised Reference Model From The Environment Helps Continual Learning0
Continual HyperTransformer: A Meta-Learner for Continual Few-Shot Learning0
Latent Spectral Regularization for Continual LearningCode0
Architect, Regularize and Replay (ARR): a Flexible Hybrid Approach for Continual Learning0
On Sequential Bayesian Inference for Continual LearningCode0
A Survey on Deep Industrial Transfer Learning in Fault Prognostics0
Generative appearance replay for continual unsupervised domain adaptation0
Dissecting Continual Learning a Structural and Data Analysis0
Dynamically Modular and Sparse General Continual LearningCode0
Self-Evolved Dynamic Expansion Model for Task-Free Continual LearningCode0
Instance and Category Supervision are Alternate Learners for Continual Learning0
Show:102550
← PrevPage 35 of 53Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1Multi-task Learning (MTL; Upper Bound)F1 - macro0.88Unverified
2CTRF1 - macro0.84Unverified
3B-CLF1 - macro0.81Unverified
4LAMOLF1 - macro0.81Unverified
5OWMF1 - macro0.79Unverified
6A-GEMF1 - macro0.78Unverified
7HATF1 - macro0.78Unverified
8Independent Learning (ONE)F1 - macro0.78Unverified
9KANF1 - macro0.77Unverified
10Naive Continual Learning (NCL)F1 - macro0.77Unverified
#ModelMetricClaimedVerifiedStatus
1NetTailordecathlon discipline (Score)3,744Unverified
2Depthwise Soft Sharingdecathlon discipline (Score)3,507Unverified
3Parallel Res. adapt.decathlon discipline (Score)3,412Unverified
4Depthwise Sharingdecathlon discipline (Score)3,234Unverified
5Series Res. adapt.decathlon discipline (Score)3,159Unverified
6Res. adapt. (large)decathlon discipline (Score)3,131Unverified
7DANdecathlon discipline (Score)2,851Unverified
8Piggybackdecathlon discipline (Score)2,838Unverified
9Res. adapt. finetune alldecathlon discipline (Score)2,643Unverified
10Res. adapt. decaydecathlon discipline (Score)2,621Unverified
#ModelMetricClaimedVerifiedStatus
1Model Zoo-ContinualAverage Accuracy94.99Unverified
2ALTAAverage Accuracy92.98Unverified
3kNN-CLIPAverage Accuracy90.8Unverified
4RMNAverage Accuracy81Unverified
5CPGAverage Accuracy80.9Unverified
6CondConvContinualAverage Accuracy77.4Unverified
7PAENetAverage Accuracy77.1Unverified
8CPG-lightAverage Accuracy77Unverified
9PackNetAverage Accuracy67.5Unverified
#ModelMetricClaimedVerifiedStatus
1ALTA-ViTB/16Average Accuracy89.8Unverified
2ALTA-RN50x4Average Accuracy84.73Unverified
3ALTA-RN101Average Accuracy83.35Unverified
4ALTA-RN50Average Accuracy81.07Unverified
5SNCLAverage Accuracy52.85Unverified
6DER [buzzega2020dark]Average Accuracy51.78Unverified
7ER[riemer2018learning]Average Accuracy48.64Unverified
8iCaRL [rebuffi2017icarl]Average Accuracy31.55Unverified
9A-GEM [chaudhry2018efficient]Average Accuracy25.33Unverified
#ModelMetricClaimedVerifiedStatus
1CAT (CNN backbone)Acc0.76Unverified
2CAT (MLP backbone)Acc0.69Unverified
3EWCAcc0.65Unverified
4HyperNetAcc0.6Unverified
5PathNetAcc0.58Unverified
6HATAcc0.57Unverified
7RPSNetAcc0.55Unverified
#ModelMetricClaimedVerifiedStatus
1CTRF1 - macro0.95Unverified
2HATF1 - macro0.95Unverified
3CATF1 - macro0.95Unverified
4B-CLF1 - macro0.95Unverified
5EWCF1 - macro0.92Unverified
6LAMOLF1 - macro0.46Unverified
#ModelMetricClaimedVerifiedStatus
1CondConvContinualAccuracy84.26Unverified
2H$^{2}$Accuracy84.1Unverified
3CPGAccuracy83.59Unverified
4PiggybackAccuracy80.5Unverified
5PackNetAccuracy80.41Unverified
6ProgressiveNetAccuracy78.94Unverified
#ModelMetricClaimedVerifiedStatus
1CTRF1 - macro0.89Unverified
2CATF1 - macro0.87Unverified
3HATF1 - macro0.86Unverified
4KANF1 - macro0.81Unverified
5B-CLF1 - macro0.77Unverified
6EWCF1 - macro0.66Unverified
#ModelMetricClaimedVerifiedStatus
1CondConvContinualAccuracy97.16Unverified
2CPGAccuracy96.62Unverified
3H$^{2}$Accuracy94.9Unverified
4PiggybackAccuracy94.77Unverified
5ProgressiveNetAccuracy93.41Unverified
6PackNetAccuracy93.04Unverified
#ModelMetricClaimedVerifiedStatus
1PiggybackAccuracy76.16Unverified
2ProgressiveNetAccuracy76.16Unverified
3CondConvContinualAccuracy76.16Unverified
4CPGAccuracy75.81Unverified
5PackNetAccuracy75.71Unverified
6H$^{2}$Accuracy75.71Unverified
#ModelMetricClaimedVerifiedStatus
1CondConvContinualAccuracy80.77Unverified
2CPGAccuracy80.33Unverified
3PiggybackAccuracy79.91Unverified
4ProgressiveNetAccuracy76.35Unverified
5H$^{2}$Accuracy76.2Unverified
6PackNetAccuracy76.17Unverified
#ModelMetricClaimedVerifiedStatus
1CPGAccuracy92.8Unverified
2CondConvContinualAccuracy92.61Unverified
3H$^{2}$Accuracy90.6Unverified
4PiggybackAccuracy89.62Unverified
5ProgressiveNetAccuracy89.21Unverified
6PackNetAccuracy86.11Unverified
#ModelMetricClaimedVerifiedStatus
1CondConvContinualAccuracy78.32Unverified
2CPGAccuracy77.15Unverified
3H$^{2}$Accuracy75.1Unverified
4ProgressiveNetAccuracy74.94Unverified
5PiggybackAccuracy71.33Unverified
6PackNetAccuracy69.4Unverified
#ModelMetricClaimedVerifiedStatus
1ALTA-ViTB/16Average Accuracy92.85Unverified
2ALTA-RN50x4Average Accuracy84.91Unverified
3RMN (Resnet)Average Accuracy84.9Unverified
4ALTA-RN101Average Accuracy84.77Unverified
5ALTA-RN50Average Accuracy83.87Unverified
#ModelMetricClaimedVerifiedStatus
1RMNAccuracy68.1Unverified
2CondConvContinualAccuracy61.32Unverified
3CCGNAccuracy35.24Unverified
4DGMwAccuracy17.82Unverified
5DGMaAccuracy15.16Unverified
#ModelMetricClaimedVerifiedStatus
1RMNAverage Accuracy97.99Unverified
2Model Zoo-ContinualAverage Accuracy97.71Unverified
3CODE-CLAverage Accuracy96.56Unverified
#ModelMetricClaimedVerifiedStatus
1TAG-RMSPropAccuracy62.59Unverified
#ModelMetricClaimedVerifiedStatus
1CODE-CLAverage Accuracy93.32Unverified
#ModelMetricClaimedVerifiedStatus
1TEST1:3 Accuracy2Unverified
#ModelMetricClaimedVerifiedStatus
1TAG-RMSPropAverage Accuracy62.79Unverified
#ModelMetricClaimedVerifiedStatus
1IBMAccuracy82.69Unverified
#ModelMetricClaimedVerifiedStatus
1IBMAccuracy88.15Unverified
#ModelMetricClaimedVerifiedStatus
1Model Zoo-ContinualAverage Accuracy84.27Unverified
#ModelMetricClaimedVerifiedStatus
1TAG-RMSPropAccuracy61.58Unverified
#ModelMetricClaimedVerifiedStatus
1CODE-CLAverage Accuracy68.83Unverified
#ModelMetricClaimedVerifiedStatus
1TAG-RMSPropAccuracy57.2Unverified
#ModelMetricClaimedVerifiedStatus
1IBMAccuracy53.9Unverified
#ModelMetricClaimedVerifiedStatus
1MRMAcc78.4Unverified
#ModelMetricClaimedVerifiedStatus
1Model Zoo-ContinualAverage Accuracy99.66Unverified
#ModelMetricClaimedVerifiedStatus
1CODE-CLAverage Accuracy77.21Unverified
#ModelMetricClaimedVerifiedStatus
1H$^{2}$Top 1 Accuracy %97.3Unverified
#ModelMetricClaimedVerifiedStatus
1H$^{2}$Top 1 Accuracy %99.9Unverified
#ModelMetricClaimedVerifiedStatus
1IBMAccuracy52.38Unverified