Multi-vision Attention Networks for On-line Red Jujube Grading

TL;DR

This paper introduces a lightweight convolutional neural network with attention mechanisms for real-time classification of red jujubes, achieving high accuracy comparable to larger models.

Contribution

The paper proposes a novel, efficient CNN model inspired by biological multi-visual mechanisms and DenseNet, enhanced with SE-Net attention for jujube grading.

Findings

Classification accuracy reaches 91.89%.

Model performs in real-time.

Comparable to larger, more complex models.

Abstract

To solve the red jujube classification problem, this paper designs a convolutional neural network model with low computational cost and high classification accuracy. The architecture of the model is inspired by the multi-visual mechanism of the organism and DenseNet. To further improve our model, we add the attention mechanism of SE-Net. We also construct a dataset which contains 23,735 red jujube images captured by a jujube grading system. According to the appearance of the jujube and the characteristics of the grading system, the dataset is divided into four classes: invalid, rotten, wizened and normal. The numerical experiments show that the classification accuracy of our model reaches to 91.89%, which is comparable to DenseNet-121, InceptionV3, InceptionV4, and Inception-ResNet v2. However, our model has real-time performance.