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 Current evaluations of Continual Learning (CL) methods typically assume that there is no constraint on training time and computation. This is an unrealistic assumption for any …
Modern machine learning pipelines are limited due to data availability, storage quotas, privacy regulations, and expensive annotation processes. These constraints make it difficult …
S Aich, J Ruiz-Santaquiteria, Z Lu… - Proceedings of the …, 2023 - openaccess.thecvf.com
This paper investigates data-free class-incremental learning (DFCIL) for hand gesture recognition from 3D skeleton sequences. In this class-incremental learning (CIL) setting …
Y Pei, Z Qing, S Zhang, X Wang… - Proceedings of the …, 2023 - openaccess.thecvf.com
Recently, prompt-based learning has made impressive progress on image class- incremental learning, but it still lacks sufficient exploration in the video domain. In this paper …
Y Pei, Z Qing, J Cen, X Wang… - Advances in …, 2022 - proceedings.neurips.cc
Recent incremental learning for action recognition usually stores representative videos to mitigate catastrophic forgetting. However, only a few bulky videos can be stored due to the …
Class-incremental learning is one of the most important settings for the study of Continual Learning, as it closely resembles real-world application scenarios. With constrained memory …
G Ding, H Golong, A Yao - … of the IEEE/CVF Conference on …, 2024 - openaccess.thecvf.com
Data replay is a successful incremental learning technique for images. It prevents catastrophic forgetting by keeping a reservoir of previous data original or synthesized to …