AutoTune: Automatically Tuning Convolutional Neural Networks for Improved Transfer Learning. (January 2021)

• Record Type: Journal Article
• Authors: Basha, S.H. Shabbeer; Vinakota, Sravan Kumar; Pulabaigari, Viswanath; Mukherjee, Snehasis; Dubey, Shiv Ram
• Is Part Of: Neural networks. Volume 133(2021)
• Page Start: 112
• Page End: 122
• Publication Date: 2021-01
• Subjects: Convolutional Neural Networks -- Fine-tuning -- Bayesian optimization -- Neural Architecture Search -- Transfer learning; Neural computers -- Periodicals; Neural networks (Computer science) -- Periodicals; Neural networks (Neurobiology) -- Periodicals; Nervous System -- Periodicals; Ordinateurs neuronaux -- Périodiques; Réseaux neuronaux (Informatique) -- Périodiques; Réseaux neuronaux (Neurobiologie) -- Périodiques; Neural computers; Neural networks (Computer science); Neural networks (Neurobiology); Periodicals; 006.32
• DOI: 10.1016/j.neunet.2020.10.009
• Languages: English
• ISSNs: 0893-6080

Abstract

Abstract: Transfer learning enables solving a specific task having limited data by using the pre-trained deep networks trained on large-scale datasets. Typically, while transferring the learned knowledge from source task to the target task, the last few layers are fine-tuned (re-trained) over the target dataset. However, these layers are originally designed for the source task that might not be suitable for the target task. In this paper, we introduce a mechanism for automatically tuning the Convolutional Neural Networks (CNN) for improved transfer learning. The pre-trained CNN layers are tuned with the knowledge from target data using Bayesian Optimization. First, we train the final layer of the base CNN model by replacing the number of neurons in the softmax layer with the number of classes involved in the target task. Next, the CNN is tuned automatically by observing the classification performance on the validation data (greedy criteria). To evaluate the performance of the proposed method, experiments are conducted on three benchmark datasets, e.g., CalTech-101, CalTech-256, and Stanford Dogs. The classification results obtained through the proposed AutoTune method outperforms the standard baseline transfer learning methods over the three datasets by achieving 95.92%, 86.54%, and 84.67% accuracy over CalTech-101, CalTech-256, and Stanford Dogs, respectively. The experimental results obtained in this study depict that tuning of the pre-trained CNN layers with the knowledge from the target dataset confesses better transfer learning ability. The source codes are available at https://github.com/JekyllAndHyde8999/AutoTune_CNN_TransferLearning . Highlights: This paper introduces a framework called AutoTune which finds the optimal number of layers to be fine-tuned automatically for a target dataset for improved transfer learning. Bayesian Optimization technique is applied to learn the pre-trained CNN layers with the knowledge of the target dataset. Several experiments are performed on CalTech-101, CalTech-256, and Stanford Dogs using the proposed AutoTune to establish the efficacy of the proposed model compared to state-of-the-art. In order to validate the robustness of the proposed framework across different CNN architectures, we experiment with three popular CNNs, namely, VGG-16 (Simonyan and Zisserman, 2014), ResNet-50 (He et al., 2016), and DenseNet-121 (Huang et al., 2017).