Q Hu, Y Gao, B Cao - … Transactions on Circuits and Systems for …, 2022 - ieeexplore.ieee.org
Modern artificial intelligence systems require class-incremental learning while suffering from catastrophic forgetting in many real-world applications. Due to the missing knowledge of …
Y Shi, D Shi, Z Qiao, Z Wang, Y Zhang, S Yang, C Qiu - Neural Networks, 2023 - Elsevier
Deep neural networks (DNNs) are prone to the notorious catastrophic forgetting problem when learning new tasks incrementally. Class-incremental learning (CIL) is a promising …
J Dong, W Liang, Y Cong… - Proceedings of the IEEE …, 2023 - openaccess.thecvf.com
Class-incremental learning (CIL) has achieved remarkable successes in learning new classes consecutively while overcoming catastrophic forgetting on old categories. However …
L Guo, G Xie, X Xu, J Ren - IEEE Access, 2020 - ieeexplore.ieee.org
Catastrophic forgetting is a key challenge for class-incremental learning with deep neural networks, where the performance decreases considerably while dealing with long …
A well-known issue for class-incremental learning is the catastrophic forgetting phenomenon, where the network's recognition performance on old classes degrades …
Abstract Class-Incremental Learning (CIL) struggles with catastrophic forgetting when learning new knowledge, and Data-Free CIL (DFCIL) is even more challenging without …
Y Choi, M El-Khamy, J Lee - Proceedings of the IEEE/CVF …, 2021 - openaccess.thecvf.com
This paper proposes two novel knowledge transfer techniques for class-incremental learning (CIL). First, we propose data-free generative replay (DF-GR) to mitigate …
Deep models, eg, CNNs and Vision Transformers, have achieved impressive achievements in many vision tasks in the closed world. However, novel classes emerge from time to time in …
Abstract Class-Incremental Learning (CIL) aims to build classification models from data streams. At each step of the CIL process, new classes must be integrated into the model …