ReFine: Re-randomization before Fine-tuning for Cross-domain Few-shot Learning

被引:4
|
作者
Oh, Jaehoon [1 ]
Kim, Sungnyun [2 ]
Ho, Namgyu [2 ]
Kim, Jin-Hwa [3 ]
Song, Hwanjun [3 ]
Yun, Se-Young [2 ]
机构
[1] KAIST DS, Daejeon, South Korea
[2] KAIST AI, Seoul, South Korea
[3] NAVER AI Lab, Sungnam, South Korea
来源
PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON INFORMATION AND KNOWLEDGE MANAGEMENT, CIKM 2022 | 2022年
关键词
cross-domain; few-shot; transfer learning; re-randomization;
D O I
10.1145/3511808.3557681
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Cross-domain few-shot learning (CD-FSL), where there are few target samples under extreme differences between source and target domains, has recently attracted huge attention. Recent studies on CD-FSL generally focus on transfer learning based approaches, where a neural network is pre-trained on popular labeled source domain datasets and then transferred to target domain data. Although the labeled datasets may provide suitable initial parameters for the target data, the domain difference between the source and target might hinder fine-tuning on the target domain. This paper proposes a simple yet powerful method that re-randomizes the parameters fitted on the source domain before adapting to the target data. The re-randomization resets source-specific parameters of the source pre-trained model and thus facilitates fine-tuning on the target domain, improving few-shot performance.
引用
收藏
页码:4359 / 4363
页数:5
相关论文
共 50 条
  • [31] Multi-level fine-tuning, data augmentation, and few-shot learning for specialized cyber threat intelligence
    Bayer, Markus
    Frey, Tobias
    Reuter, Christian
    COMPUTERS & SECURITY, 2023, 134
  • [32] Cross-Domain Few-Shot Anomaly Detection for equipment in nuclear power plants
    He, Junjie
    Zheng, Sheng
    Yi, Shuang
    Yang, Senquan
    Huan, Zhihe
    NUCLEAR ENGINEERING AND DESIGN, 2025, 436
  • [33] Learned Gaussian ProtoNet for improved cross-domain few-shot classification and generalization
    Khanday, Nadeem Yousuf
    Sofi, Shabir Ahmad
    NEURAL COMPUTING & APPLICATIONS, 2023, 35 (04) : 3435 - 3448
  • [34] Novel joint transfer fine-grained metric network for cross-domain few-shot fault diagnosis
    Hu, Junwei
    Li, Weigang
    Wu, Ailong
    Tian, Zhiqiang
    KNOWLEDGE-BASED SYSTEMS, 2023, 279
  • [35] Learned Gaussian ProtoNet for improved cross-domain few-shot classification and generalization
    Nadeem Yousuf Khanday
    Shabir Ahmad Sofi
    Neural Computing and Applications, 2023, 35 : 3435 - 3448
  • [36] Metric-Based Meta-Learning Relation Network for Cross-Domain Few-Shot Bearing Fault Diagnosis
    Gao, Wei
    Xu, Zhiqiang
    Akoudad, Youssef
    IEEE SENSORS JOURNAL, 2025, 25 (08) : 13632 - 13647
  • [37] Covariance-based metric model for cross-domain few-shot classification and learning-to-generalization
    Nadeem Yousuf Khanday
    Shabir Ahmad Sofi
    Applied Intelligence, 2023, 53 : 27374 - 27391
  • [38] Covariance-based metric model for cross-domain few-shot classification and learning-to-generalization
    Khanday, Nadeem Yousuf
    Sofi, Shabir Ahmad
    APPLIED INTELLIGENCE, 2023, 53 (22) : 27374 - 27391
  • [39] Cross-Domain Meta-Learning Under Dual-Adjustment Mode for Few-Shot Hyperspectral Image Classification
    Hu, Lei
    He, Wei
    Zhang, Liangpei
    Zhang, Hongyan
    IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2023, 61
  • [40] User-Specific Adaptive Fine-Tuning for Cross-Domain Recommendations
    Chen, Lei
    Yuan, Fajie
    Yang, Jiaxi
    He, Xiangnan
    Li, Chengming
    Yang, Min
    IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, 2023, 35 (03) : 3239 - 3252