CVPR 2021 accepted papers on Continual learning
25 of the 1,661 papers accepted at CVPR 2021 (1.5%) are on Continual learning, compared with 0.9% at CVPR 2020. They fall into 3 topics.
See CVPR 2021 in the interactive map
Submitting to CVPR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Class-incremental · 12 papers
- DER: Dynamically Expandable Representation for Class Incremental Learning (653 citations)
- Prototype Augmentation and Self-Supervision for Incremental Learning (403 citations)
- Few-Shot Incremental Learning With Continually Evolved Classifiers (363 citations)
- PLOP: Learning Without Forgetting for Continual Semantic Segmentation (265 citations)
- Adaptive Aggregation Networks for Class-Incremental Learning (227 citations)
- Semantic-Aware Knowledge Distillation for Few-Shot Class-Incremental Learning (192 citations)
- Self-Promoted Prototype Refinement for Few-Shot Class-Incremental Learning (183 citations)
- Distilling Causal Effect of Data in Class-Incremental Learning (167 citations)
- Continual Semantic Segmentation via Repulsion-Attraction of Sparse and Disentangled Latent Representations (152 citations)
- Efficient Feature Transformations for Discriminative and Generative Continual Learning (62 citations)
- Rectification-Based Knowledge Retention for Continual Learning (45 citations)
- Incremental Learning via Rate Reduction (37 citations)
Online and replay-based · 11 papers
- Rainbow Memory: Continual Learning With a Memory of Diverse Samples (347 citations)
- Mixed-Privacy Forgetting in Deep Networks (123 citations)
- On Learning the Geodesic Path for Incremental Learning (123 citations)
- Training Networks in Null Space of Feature Covariance for Continual Learning (116 citations)
- ORDisCo: Effective and Efficient Usage of Incremental Unlabeled Data for Semi-Supervised Continual Learning (74 citations)
- Layerwise Optimization by Gradient Decomposition for Continual Learning (62 citations)
- Image De-Raining via Continual Learning (50 citations)
- Continual Learning via Bit-Level Information Preserving (36 citations)
- Hyper-LifelongGAN: Scalable Lifelong Learning for Image Conditioned Generation (36 citations)
- IIRC: Incremental Implicitly-Refined Classification (23 citations)
- KSM: Fast Multiple Task Adaption via Kernel-Wise Soft Mask Learning (11 citations)
Domain and prompt-based · 2 papers
- Lifelong Person Re-Identification via Adaptive Knowledge Accumulation (99 citations)
- Continual Adaptation of Visual Representations via Domain Randomization and Meta-Learning (74 citations)
