Using low-rank adaptation to fine-tune pre-trained neural network models has attracted widespread attention due to its advantages of low resource requirements, high precision, and no additional inference delay. However, most existing methods are designed for large language models based on the Transformer structure and lack adaptation to convolutional neural networks (CNN) widely used on Internet of Things (IoT) devices. In addition, IoT devices are usually deployed outdoors and collect large amounts of data that are affected by environmental conditions. Providing a highly robust CNN model is a prerequisite for providing high-quality services. To this end, this paper proposes a highly robust seed-filter adaptation method (SeFA) for pre-trained CNNs. SeFA introduces an adaptation branch with the same structure as the backbone network. In the adaptation branch, some filters are first designated seed filters and grouped. Then, additional filters are generated from the grouped seed filters and nonlinear transformation functions (NLFs) with different hyperparameters. The seed filters' parameters are updated during model training, and the NLF hyperparameters are randomly initialized and frozen. Both grouping seed filters and configuring NLFs with non-learnable hyperparameters can improve the model's robustness. This is because grouping seed filters can generate diverse filters on demand without increasing the model's complexity, and the NLFs' rules can regularize the model. It is worth mentioning that the number of fine-tuning parameters of the pre-trained model that can adapt to downstream tasks can be flexibly controlled by specifying the number of seed filters. The key idea of this paper is to propose SeFA with flexible, controllable, learnable parameters, high robustness, and adaptability to pretrained CNN models, thereby facilitating fine-tuning on resource-constrained IoT devices and providing highly robust visual services. Experimental results on the CIFAR-10, CIFAR-10-C, CIFAR-100, CIFAR-100-C, and Icons-50 datasets demonstrate that the proposed SeFA outperforms other state-of-the-art methods. Specifically, on the ResNet-152 model and the CIFAR-10-C dataset, the accuracy of our SeFA is about 7% higher than that of the full fine-tuning method. In addition, we verify the efficiency of SeFA-based models for fine-tuning and inference on the device.

Using low-rank adaptation to fine-tuning pretrained neural network models has attracted widespread attention due to its advantages of low resource requirements, high precision, and no additional inference delay. However, most existing methods are designed for large language models based on Transformer structure and lack adaptation to convolutional neural networks (CNN) widely used on Internet of Things (IoT) devices. In addition, IoT devices are usually deployed outdoors and collect a large amount of data affected by the environment. Providing a highly robust model is a prerequisite for providing high-quality services. To this end, this paper proposes a highly robust seed-filter adaptation method (SeFA) for pre-trained CNNs. SeFA introduces an adaptation branch with the same structure as the backbone network. In the adaptation branch, some filters are first designated seed filters and grouped. Then, other filters are generated based on the grouped seed filters and nonlinear transformation functions (NLFs) with different hyperparameters. The parameters of the seed filters are updated with model training, and the hyperparameters of the NLFs are randomly initialized and frozen. Both grouping seed filters and configuring NLFs with nonlearnable hyperparameters can improve the robustness of the model. This is because grouping seed filters can generate diverse filters on demand without increasing the model's complexity, and the NLFs' rules can regularize the model. The key idea of this paper is to propose SeFA with flexible controllable learnable parameters, high robustness and adaptability to pretrained CNN models, to facilitate fine-tuning of pre-trained CNN models on resource-constrained IoT devices to provide highly robust visual services. Experimental results on the CIFAR-10, CIFAR-10-C, CIFAR-100, CIFAR-100-C, and Icons50 datasets demonstrate that the proposed SeFA outperforms other state-of-the-art methods. Specifically, based on the ResNet152, on the CIFAR-10-C dataset, the accuracy of our SeFA is about$+7 {\%}$higher than that of the full fine-tuning method.