Analysis on DeepLabV3+ Performance for Automatic Steel Defects Detection
This work addresses steel defect detection for industrial quality control, but is incremental as it applies existing methods to a specific domain.
The researchers tackled the problem of automatically detecting steel defects by experimenting with DeepLabV3+ with different backbones on a large dataset of steel images, finding that ResNet101 and EfficientNet backbones achieved the best IoU score of around 0.57 compared to 0.325 for DenseNet.
Our works experimented DeepLabV3+ with different backbones on a large volume of steel images aiming to automatically detect different types of steel defects. Our methods applied random weighted augmentation to balance different defects types in the training set. And then applied DeeplabV3+ model three different backbones, ResNet, DenseNet and EfficientNet, on segmenting defection regions on the steel images. Based on experiments, we found that applying ResNet101 or EfficientNet as backbones could reach the best IoU scores on the test set, which is around 0.57, comparing with 0.325 for using DenseNet. Also, DeepLabV3+ model with ResNet101 as backbone has the fewest training time.