Rock Classification Based on Residual Networks

TL;DR

This work addresses rock classification under limited data by exploring residual-network–based approaches. It combines data augmentation with a ResNet34 backbone, introduces ConvNeXt-inspired kernel modifications, and integrates a Bottleneck Transformer with internal residual connections to leverage attention within a CNN framework. Empirical results show data augmentation raising accuracy to $66.64\%$, kernel modifications to $70.1\%$, and a BoTNet with IRC achieving the top $73.7\%$ on the test set, while more than one transformer block can degrade performance. The findings offer practical guidance for designing residual architectures under data-scarce conditions and motivate future work on larger datasets and joint application of the proposed techniques.

Abstract

Rock Classification is an essential geological problem since it provides important formation information. However, exploration on this problem using convolutional neural networks is not sufficient. To tackle this problem, we propose two approaches using residual neural networks. We first adopt data augmentation methods to enlarge our dataset. By modifying kernel sizes, normalization methods and composition based on ResNet34, we achieve an accuracy of 70.1% on the test dataset, with an increase of 3.5% compared to regular Resnet34. Furthermore, using a similar backbone like BoTNet that incorporates multihead self attention, we additionally use internal residual connections in our model. This boosts the model's performance, achieving an accuracy of 73.7% on the test dataset. We also explore how the number of bottleneck transformer blocks may influence model performance. We discover that models with more than one bottleneck transformer block may not further improve performance. Finally, we believe that our approach can inspire future work related to this problem and our model design can facilitate the development of new residual model architectures.

Figures (5)

• Figure 1: The data augmentation process
• Figure 2: Some examples randomly extracted after data augmentation
• Figure 3: Layout of a bottleneck transformer block. Left: A normal bottleneck transformer block. Right: A bottleneck transformer block with internal residual connection.
• Figure 4: The structure of the modified kernel.
• Figure 5: Layouts of the last convolution layer. Left: Original ResNet34 layout. Middle: One bottleneck transformer block. Right: Two bottleneck transformer blocks.