Parameter-Efficient Multi-Task and Multi-Domain Learning Using Factorized Tensor Networks
Open Access
- 22 September 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Signal Processing
- Vol. 6, 1077-1085
Abstract
Multi-task and multi-domain learning methods seek to learn multiple tasks/domains, jointly or one after another, using a single unified network. The primary challenge and opportunity lie in leveraging shared information across these tasks and domains to enhance the efficiency of the unified network. The efficiency can be in terms of accuracy, storage cost, computation, or sample complexity. In this paper, we introduce a factorized tensor network (FTN) designed to achieve accuracy comparable to that of independent single-task or single-domain networks, while introducing a minimal number of additional parameters. The FTN approach entails incorporating task- or domain-specific low-rank tensor factors into a shared frozen network derived from a source model. This strategy allows for adaptation to numerous target domains and tasks without encountering catastrophic forgetting. Furthermore, FTN requires a significantly smaller number of task-specific parameters compared to existing methods. We performed experiments on widely used multi-domain and multi-task datasets. We show the experiments on convolutional-based architecture with different backbones and on transformer-based architecture. Our findings indicate that FTN attains similar accuracy as single-task or single-domain methods while using only a fraction of additional parameters per task.Keywords
All Related Versions
Funding Information
- NSF CAREER (CCF-2046293)
- Air Force Office of Scientific Research (FA9550-21-1-0330)
- ONR (N00014-19-1-2264)
- USDA (2023-67021-40629)
This publication has 26 references indexed in Scilit:
- Inverted Pyramid Multi-task Transformer for Dense Scene Understanding2022
- Factorizing Knowledge in Neural Networks2022
- Side-Tuning: A Baseline for Network Adaptation via Additive Side Networks2020
- Reparameterizing Convolutions for Incremental Multi-Task Learning Without Task Interference2020
- Adding New Tasks to a Single Network with Weight Transformations Using Binary Masks2019
- Piggyback: Adapting a Single Network to Multiple Tasks by Learning to Mask Weights2018
- Deep Hashing Network for Unsupervised Domain Adaptation2017
- Adversarial Discriminative Domain Adaptation2017
- Cross-Stitch Networks for Multi-task Learning2016
- Adapting Visual Category Models to New Domains2010