File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1007/978-3-031-19812-0_18
- Scopus: eid_2-s2.0-85142734874
- WOS: WOS:000903590200018
- Find via
Supplementary
- Citations:
- Appears in Collections:
Conference Paper: MTFormer: Multi-task Learning via Transformer and Cross-Task Reasoning
Title | MTFormer: Multi-task Learning via Transformer and Cross-Task Reasoning |
---|---|
Authors | |
Keywords | Cross-task reasoning Multi-task learning Transformer |
Issue Date | 2022 |
Citation | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2022, v. 13687 LNCS, p. 304-321 How to Cite? |
Abstract | In this paper, we explore the advantages of utilizing transformer structures for addressing multi-task learning (MTL). Specifically, we demonstrate that models with transformer structures are more appropriate for MTL than convolutional neural networks (CNNs), and we propose a novel transformer-based architecture named MTFormer for MTL. In the framework, multiple tasks share the same transformer encoder and transformer decoder, and lightweight branches are introduced to harvest task-specific outputs, which increases the MTL performance and reduces the time-space complexity. Furthermore, information from different task domains can benefit each other, and we conduct cross-task reasoning. We propose a cross-task attention mechanism for further boosting the MTL results. The cross-task attention mechanism brings little parameters and computations while introducing extra performance improvements. Besides, we design a self-supervised cross-task contrastive learning algorithm for further boosting the MTL performance. Extensive experiments are conducted on two multi-task learning datasets, on which MTFormer achieves state-of-the-art results with limited network parameters and computations. It also demonstrates significant superiorities for few-shot learning and zero-shot learning. |
Persistent Identifier | http://hdl.handle.net/10722/333567 |
ISSN | 2023 SCImago Journal Rankings: 0.606 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Xu, Xiaogang | - |
dc.contributor.author | Zhao, Hengshuang | - |
dc.contributor.author | Vineet, Vibhav | - |
dc.contributor.author | Lim, Ser Nam | - |
dc.contributor.author | Torralba, Antonio | - |
dc.date.accessioned | 2023-10-06T05:20:38Z | - |
dc.date.available | 2023-10-06T05:20:38Z | - |
dc.date.issued | 2022 | - |
dc.identifier.citation | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2022, v. 13687 LNCS, p. 304-321 | - |
dc.identifier.issn | 0302-9743 | - |
dc.identifier.uri | http://hdl.handle.net/10722/333567 | - |
dc.description.abstract | In this paper, we explore the advantages of utilizing transformer structures for addressing multi-task learning (MTL). Specifically, we demonstrate that models with transformer structures are more appropriate for MTL than convolutional neural networks (CNNs), and we propose a novel transformer-based architecture named MTFormer for MTL. In the framework, multiple tasks share the same transformer encoder and transformer decoder, and lightweight branches are introduced to harvest task-specific outputs, which increases the MTL performance and reduces the time-space complexity. Furthermore, information from different task domains can benefit each other, and we conduct cross-task reasoning. We propose a cross-task attention mechanism for further boosting the MTL results. The cross-task attention mechanism brings little parameters and computations while introducing extra performance improvements. Besides, we design a self-supervised cross-task contrastive learning algorithm for further boosting the MTL performance. Extensive experiments are conducted on two multi-task learning datasets, on which MTFormer achieves state-of-the-art results with limited network parameters and computations. It also demonstrates significant superiorities for few-shot learning and zero-shot learning. | - |
dc.language | eng | - |
dc.relation.ispartof | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | - |
dc.subject | Cross-task reasoning | - |
dc.subject | Multi-task learning | - |
dc.subject | Transformer | - |
dc.title | MTFormer: Multi-task Learning via Transformer and Cross-Task Reasoning | - |
dc.type | Conference_Paper | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1007/978-3-031-19812-0_18 | - |
dc.identifier.scopus | eid_2-s2.0-85142734874 | - |
dc.identifier.volume | 13687 LNCS | - |
dc.identifier.spage | 304 | - |
dc.identifier.epage | 321 | - |
dc.identifier.eissn | 1611-3349 | - |
dc.identifier.isi | WOS:000903590200018 | - |