TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Papers/iCaRL: Incremental Classifier and Representation Learning

iCaRL: Incremental Classifier and Representation Learning

Sylvestre-Alvise Rebuffi, Alexander Kolesnikov, Georg Sperl, Christoph H. Lampert

2016-11-23CVPR 2017 7Representation LearningClass Incremental Learningclass-incremental learningIncremental Learning
PaperPDFCode(official)CodeCodeCodeCodeCodeCodeCodeCodeCode

Abstract

A major open problem on the road to artificial intelligence is the development of incrementally learning systems that learn about more and more concepts over time from a stream of data. In this work, we introduce a new training strategy, iCaRL, that allows learning in such a class-incremental way: only the training data for a small number of classes has to be present at the same time and new classes can be added progressively. iCaRL learns strong classifiers and a data representation simultaneously. This distinguishes it from earlier works that were fundamentally limited to fixed data representations and therefore incompatible with deep learning architectures. We show by experiments on CIFAR-100 and ImageNet ILSVRC 2012 data that iCaRL can learn many classes incrementally over a long period of time where other strategies quickly fail.

Results

TaskDatasetMetricValueModel
Continual Learningcifar10010-stage average accuracy63.24iCaRL
Incremental LearningCIFAR-100 - 50 classes + 10 steps of 5 classesAverage Incremental Accuracy52.57iCaRL*
Incremental LearningCIFAR-100 - 50 classes + 5 steps of 10 classesAverage Incremental Accuracy57.17iCaRL*
Incremental LearningCIFAR-100-B0(5steps of 20 classes)Average Incremental Accuracy71.14iCaRL
Incremental LearningImageNet - 10 steps# M Params11.68iCaRL
Incremental LearningImageNet - 10 stepsAverage Incremental Accuracy38.4iCaRL
Incremental LearningImageNet - 10 stepsAverage Incremental Accuracy Top-563.7iCaRL
Incremental LearningImageNet - 10 stepsFinal Accuracy22.7iCaRL
Incremental LearningImageNet - 10 stepsFinal Accuracy Top-544iCaRL
Incremental LearningImageNet100 - 10 steps# M Params11.22iCaRL
Incremental LearningImageNet100 - 10 stepsAverage Incremental Accuracy Top-583.6iCaRL
Incremental LearningImageNet100 - 10 stepsFinal Accuracy Top-563.8iCaRL
Incremental LearningCIFAR-100 - 50 classes + 2 steps of 25 classesAverage Incremental Accuracy71.33iCaRL
Incremental LearningImageNet-100 - 50 classes + 5 steps of 10 classesAverage Incremental Accuracy65.56iCaRL*
Class Incremental Learningcifar10010-stage average accuracy63.24iCaRL

Related Papers

Touch in the Wild: Learning Fine-Grained Manipulation with a Portable Visuo-Tactile Gripper2025-07-20Spectral Bellman Method: Unifying Representation and Exploration in RL2025-07-17Boosting Team Modeling through Tempo-Relational Representation Learning2025-07-17Similarity-Guided Diffusion for Contrastive Sequential Recommendation2025-07-16Are encoders able to learn landmarkers for warm-starting of Hyperparameter Optimization?2025-07-16Language-Guided Contrastive Audio-Visual Masked Autoencoder with Automatically Generated Audio-Visual-Text Triplets from Videos2025-07-16A Mixed-Primitive-based Gaussian Splatting Method for Surface Reconstruction2025-07-15Dual Dimensions Geometric Representation Learning Based Document Dewarping2025-07-11