Image Classification
Introduction
Image classification is used to identify and distinguish the overall content of images, and is one of the most fundamental problems in computer vision. For image classification problems, we provide an automated modeling solution that enables users to solve image classification problems they encounter on their own.
Evaluation Metrics
Top-1 Accuracy
Top-5 Accuracy
Parameter Explanation
| Parameter | Description |
|---|---|
| image_height | (Required) Image height |
| image_width | (Required) Image width |
| train_meta | (Required) Training set CSV file path |
| val_meta | (Required) Validation set CSV file path |
| test_meta | (Optional) Test set CSV file path |
Practical Examples
In this section, we will demonstrate the results on several example problems in combination with the data format.
Surface Crack Image Classification
Original problem link: https://www.kaggle.com/arunrk7/surface-crack-detection This problem involves detecting defects on concrete surfaces, where cracks indicate defects. The dataset used for this problem contains 20,000 images each for positive and negative examples, with the training and validation sets randomly sampled at an 8:2 ratio.
The model obtained through automated modeling achieved an accuracy of 99.775%.