.png){kind=icon}
.png){kind=icon}
Return
Model adaptive parameter fine-tuning Based on contribution measure for image classification
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2025.129634.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Fine-tuning is an important transfer learning technique that has achieved significant success in various image classification tasks lacking training data and requires only a small number of training epochs to achieve satisfactory results. However, with the increasing complexity of the model scale and structure, designing appropriate fine-tuning schemes for specific target tasks becomes increasingly difficult. In this paper, a contribution measure criterion is used to quantify the importance of the pre-trained model parameters to the target task, providing a basis for selecting fine-tuning parameters. In addition, we find that the fine-tuning ratio vary depends on the specific target task. Therefore, we propose an adaptive fine-tuning ratio search strategy to search the appropriate fine-tuning ratio for the given target task. Based on the above strategy, we propose an adaptive fine-tuning algorithm based on parameter contribution to customize the fine-tuning scheme for the target task. The experimental results show that the proposed algorithm can effectively quantify the contribution of model parameters, and our algorithm can adaptively adjust the fine-tuning ratio for the target task. Furthermore, our algorithm achieves state-of-the-art performance on seven publicly available image classification datasets widely used in transfer learning.
Keywords:
Parameter fine-tuning
Transfer learning
Image classification
Contribution measure
Fine-tuning ratio