Knowledge Commons of Institute of Automation,CAS
Pro-tuning: Unified Prompt Tuning for Vision Tasks | |
Xing Nie1,2![]() ![]() ![]() ![]() ![]() | |
发表期刊 | IEEE Transactions on Circuits and Systems for Video Technology
![]() |
2023-10 | |
卷号 | 34期号:6页码:4653 - 4667 |
摘要 | In computer vision, fine-tuning is the de-facto approach to leverage pre-trained vision models to perform downstream tasks. However, deploying it in practice is quite challenging, due to adopting parameter inefficient global update and heavily relying on high-quality downstream data. Recently, prompt-based learning, which adds the task-relevant prompt to adapt the pre-trained models to downstream tasks, has drastically boosted the performance of many natural language downstream tasks. In this work, we extend this notable transfer ability benefited from prompt into vision models as an alternative to fine-tuning. To this end, we propose parameter-efficient Prompt tuning (Pro-tuning) to adapt diverse frozen pre-trained models to a wide variety of downstream vision tasks. The key to Pro-tuning is prompt-based tuning, i.e., learning task-specific vision prompts for downstream input images with the pre-trained model frozen. By only training a small number of additional parameters, Protuning can generate compact and robust downstream models both for CNN-based and transformer-based network architectures. Comprehensive experiments evidence that the proposed Protuning outperforms fine-tuning on a broad range of vision tasks and scenarios, including image classification (under generic objects, class imbalance, image corruption, adversarial robustness, and out-of-distribution generalization), and dense prediction tasks such as object detection and semantic segmentation. |
七大方向——子方向分类 | 图像视频处理与分析 |
国重实验室规划方向分类 | 环境多维感知 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 期刊论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/57463 |
专题 | 多模态人工智能系统全国重点实验室_先进时空数据分析与学习 |
作者单位 | 1.the Department of State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Science 2.the School of Artificial Intelligence, University of Chinese Academy of Sciences 3.CAS Centre for Artificial Intelligence and Robotics, HK Institute of Science and Innovation 4.Huawei Cloud & AI |
推荐引用方式 GB/T 7714 | Xing Nie,Bolin Ni,Jianlong Chang,et al. Pro-tuning: Unified Prompt Tuning for Vision Tasks[J]. IEEE Transactions on Circuits and Systems for Video Technology,2023,34(6):4653 - 4667. |
APA | Xing Nie.,Bolin Ni.,Jianlong Chang.,Gaofeng Meng.,Chunlei Huo.,...&Qi Tian.(2023).Pro-tuning: Unified Prompt Tuning for Vision Tasks.IEEE Transactions on Circuits and Systems for Video Technology,34(6),4653 - 4667. |
MLA | Xing Nie,et al."Pro-tuning: Unified Prompt Tuning for Vision Tasks".IEEE Transactions on Circuits and Systems for Video Technology 34.6(2023):4653 - 4667. |
条目包含的文件 | ||||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
Pro-tuning_Unified_P(2224KB) | 期刊论文 | 作者接受稿 | 开放获取 | CC BY-NC-SA | 浏览 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论