CLoRA: Parameter-Efficient Continual Learning with Low-Rank Adaptation

Shishir Muralidhara, Didier Stricker, René Schuster
Proceedings of The 4th Conference on Lifelong Learning Agents, PMLR 330:696-712, 2026.

Abstract

In the past, continual learning (CL) was mostly concerned with the problem of catastrophic forgetting in neural networks, that arises when incrementally learning a sequence of tasks. Current CL methods function within the confines of limited data access, without any restrictions imposed on computational resources. However, in real-world scenarios, the latter takes precedence as deployed systems are often computationally constrained. A major drawback of most CL methods is the need to retrain the entire model for each new task. The computational demands of retraining large models can be prohibitive, limiting the applicability of CL in environments with limited resources. Through CLoRA, we explore the applicability of Low-Rank Adaptation (LoRA), a parameter-efficient fine-tuning method for class-incremental semantic segmentation. CLoRA leverages a small set of parameters of the model and uses the same set for learning across all tasks. Results demonstrate the efficacy of CLoRA, achieving performance on par with and exceeding the baseline methods. We further evaluate CLoRA using NetScore, underscoring the need to factor in resource efficiency and evaluate CL methods beyond task performance. CLoRA significantly reduces the hardware requirements for training, making it well-suited for CL in resource-constrained environments after deployment.

Cite this Paper

BibTeX
@InProceedings{pmlr-v330-muralidhara26a, title = {CLoRA: Parameter-Efficient Continual Learning with Low-Rank Adaptation}, author = {Muralidhara, Shishir and Stricker, Didier and Schuster, Ren\'{e}}, booktitle = {Proceedings of The 4th Conference on Lifelong Learning Agents}, pages = {696--712}, year = {2026}, editor = {Chandar, Sarath and Pascanu, Razvan and Eaton, Eric and Liu, Bing and Mahmood, Rupam and Rannen-Triki, Amal}, volume = {330}, series = {Proceedings of Machine Learning Research}, month = {11--14 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v330/main/assets/muralidhara26a/muralidhara26a.pdf}, url = {https://proceedings.mlr.press/v330/muralidhara26a.html}, abstract = {In the past, continual learning (CL) was mostly concerned with the problem of catastrophic forgetting in neural networks, that arises when incrementally learning a sequence of tasks. Current CL methods function within the confines of limited data access, without any restrictions imposed on computational resources. However, in real-world scenarios, the latter takes precedence as deployed systems are often computationally constrained. A major drawback of most CL methods is the need to retrain the entire model for each new task. The computational demands of retraining large models can be prohibitive, limiting the applicability of CL in environments with limited resources. Through CLoRA, we explore the applicability of Low-Rank Adaptation (LoRA), a parameter-efficient fine-tuning method for class-incremental semantic segmentation. CLoRA leverages a small set of parameters of the model and uses the same set for learning across all tasks. Results demonstrate the efficacy of CLoRA, achieving performance on par with and exceeding the baseline methods. We further evaluate CLoRA using NetScore, underscoring the need to factor in resource efficiency and evaluate CL methods beyond task performance. CLoRA significantly reduces the hardware requirements for training, making it well-suited for CL in resource-constrained environments after deployment.} }
Endnote
%0 Conference Paper %T CLoRA: Parameter-Efficient Continual Learning with Low-Rank Adaptation %A Shishir Muralidhara %A Didier Stricker %A René Schuster %B Proceedings of The 4th Conference on Lifelong Learning Agents %C Proceedings of Machine Learning Research %D 2026 %E Sarath Chandar %E Razvan Pascanu %E Eric Eaton %E Bing Liu %E Rupam Mahmood %E Amal Rannen-Triki %F pmlr-v330-muralidhara26a %I PMLR %P 696--712 %U https://proceedings.mlr.press/v330/muralidhara26a.html %V 330 %X In the past, continual learning (CL) was mostly concerned with the problem of catastrophic forgetting in neural networks, that arises when incrementally learning a sequence of tasks. Current CL methods function within the confines of limited data access, without any restrictions imposed on computational resources. However, in real-world scenarios, the latter takes precedence as deployed systems are often computationally constrained. A major drawback of most CL methods is the need to retrain the entire model for each new task. The computational demands of retraining large models can be prohibitive, limiting the applicability of CL in environments with limited resources. Through CLoRA, we explore the applicability of Low-Rank Adaptation (LoRA), a parameter-efficient fine-tuning method for class-incremental semantic segmentation. CLoRA leverages a small set of parameters of the model and uses the same set for learning across all tasks. Results demonstrate the efficacy of CLoRA, achieving performance on par with and exceeding the baseline methods. We further evaluate CLoRA using NetScore, underscoring the need to factor in resource efficiency and evaluate CL methods beyond task performance. CLoRA significantly reduces the hardware requirements for training, making it well-suited for CL in resource-constrained environments after deployment.
APA
Muralidhara, S., Stricker, D. & Schuster, R.. (2026). CLoRA: Parameter-Efficient Continual Learning with Low-Rank Adaptation. Proceedings of The 4th Conference on Lifelong Learning Agents, in Proceedings of Machine Learning Research 330:696-712 Available from https://proceedings.mlr.press/v330/muralidhara26a.html.

Related Material