Abstract:In this study, the optimization of a Convolutional Neural Network (CNN) was conducted using the Fruits 360 dataset, with a specific emphasis on the impacts of Spatial Transformer Network (STN) and Stochastic Gradient Descent (SGD) optimization methods. Firstly, a baseline CNN model is built, which achieves 97.84% accuracy with a loss of 0.0999 after 50 epochs. Then, the impact of integrating STN and SGD into CNN models separately is investigated. The addition of STN slightly increased the accuracy to 97.92%, reduced the loss to 0.0994, and decreased the validation accuracy. This result suggests that while STN enhances the model's generalization ability, it may slightly reduce the maximum accuracy achievable on the validation set. After SGD optimization, the verification accuracy is increased to 98.19%, the loss is reduced to 0.0537, and the verification accuracy is increased to 98.40%. These results highlight the effectiveness of SGD in fine-tuning model parameters, resulting in more accurate models and improved generalization capabilities. A comparative analysis of these methods highlights their respective advantages. The effectiveness of the STN is rooted in its capacity to improve model generalization and mitigate overfitting, which is particularly beneficial in situations that demand robustness against varied data sets. In contrast, SGD stands out for its ability to significantly improve model accuracy and reduce loss, making it a balanced choice for comprehensive model optimization. Future research directions include exploring these optimization techniques on various datasets and investigating the potential of combining STN and SGD to achieve higher performance in CNN models.

Improved Method of Convolution Neural Network Based on Matrix Decomposition

Advances in Convolutional Neural Networks

Convolutional neural networks with low-rank regularization

Image Target Recognition Based on Improved Convolutional Neural Network

Improving Efficiency in Convolutional Neural Network with Multilinear Filters

Nonlinear CNN: Improving CNNs with Quadratic Convolutions

Speeding Up Deep Convolutional Neural Networks Based on Tucker-CP Decomposition

Towards Better Analysis of Deep Convolutional Neural Networks

Depth-wise Decomposition for Accelerating Separable Convolutions in Efficient Convolutional Neural Networks

Applying Improved Convolutional Neural Network in Image Classification

IC Networks: Remodeling the Basic Unit for Convolutional Neural Networks

Optimization of CNN through Novel Training Strategy for Visual Classification Problems

Deep CNNs Meet Global Covariance Pooling: Better Representation and Generalization

Speeding-up and compression convolutional neural networks by low-rank decomposition without fine-tuning

Tensor Decomposition for Model Reduction in Neural Networks: A Review

Layer-Specific Optimization: Sensitivity Based Convolution Layers Basis Search

The Investigation of Multiple Optimization Methods on Convolutional Neural Network

Convolutional Neural Network Compression Based on Low-Rank Decomposition

Accelerating Very Deep Convolutional Networks for Classification and Detection

Enhancing Convolutional Neural Networks with Higher-Order Numerical Difference Methods

Structured Convolutions for Efficient Neural Network Design