Well Begun is Half Done: Efficient Initialization for Online Continual Learning via Two-Stage Model Merging
Open Access
- 9 November 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 6291-6300
Abstract
No abstract availableThis publication has 17 references indexed in Scilit:
- A Review of Continual Learning in Edge AIIEEE Transactions on Network Science and Engineering, 2026
- Task Singular Vectors: Reducing Task Interference in Model Merging2025
- Evolutionary optimization of model merging recipesNature Machine Intelligence, 2025
- Loss of plasticity in deep continual learningNature, 2024
- Few-Shot Class-Incremental Learning via Class-Aware Bilateral Distillation2023
- Dynamic Support Network for Few-Shot Class Incremental LearningIEEE Transactions on Pattern Analysis and Machine Intelligence, 2022
- Smart surveillance system for real-time multi-person multi-camera tracking at the edgeJournal of Real-Time Image Processing, 2021
- Deep CNN-Based Real-Time Traffic Light Detector for Self-Driving VehiclesIEEE Transactions on Mobile Computing, 2020
- Development of UAV-Based Target Tracking and Recognition SystemsIEEE Transactions on Intelligent Transportation Systems, 2019
- Warm starting the projected Gauss–Seidel algorithm for granular matter simulationComputational Particle Mechanics, 2016