Abstract
With ever-increasing data volumes, it is essential to develop automated approaches for identifying nanoscale defects in transmission electron microscopy (TEM) images. However, compared to features in conventional photographs, nanoscale defects in TEM images exhibit far greater variation due to the complex contrast mechanisms and intricate defect structures. These challenges often result in much less labeled data and higher rates of annotation errors, posing significant obstacles to improving machine learning model performance for TEM image analysis. To address these limitations, we examined transfer learning by leveraging large, pre-trained models used for natural images. We demonstrated that by using the pre-trained encoder and L2-regularization, semantically complex features are ignored in favor of simpler, more reliable cues, substantially improving the model performance. However, this improvement cannot be captured by conventional evaluation metrics such as F1-score, which can be skewed by human annotation errors treated as ground truth. Instead, we introduced novel evaluation metrics that are independent of the annotation accuracy. Using grain boundary detection in UO2 TEM images as a case study, we found that our approach led to a 57% improvement in defect detection rate, which is a robust and holistic measure of model performance on the TEM dataset used in this work. Finally, we showed that model self-confidence is only achieved through transfer learning and fine-tuning of very deep layers.
Abstract (translated)
随着数据量的不断增加，开发自动识别透射电子显微镜（TEM）图像中纳米级缺陷的方法变得至关重要。然而，与传统照片中的特征相比，由于复杂的对比机制和复杂的缺陷结构，TEM图像中的纳米级缺陷表现出更大的变异性。这些挑战通常导致标记的数据大大减少，并且注释错误率更高，这对提高机器学习模型在TEM图像分析方面的性能构成了重大障碍。 为了解决这些问题，我们研究了迁移学习方法，通过利用大型预训练模型来处理自然图像。我们证明，在使用预训练编码器和L2正则化的情况下，复杂的语义特征会被忽略，而更简单且可靠的线索则被保留下来，从而显著提高了模型的性能。然而，这种改进无法通过传统的评估指标（如F1分数）捕捉到，因为这些传统指标可能会受到被视为真实标签的人类注释错误的影响。因此，我们引入了独立于标注准确性的新型评价指标。 以UO2 TEM图像中的晶界检测为例，我们的方法将缺陷检出率提高了57%，这是一个在用于本工作的TEM数据集上衡量模型性能的稳健且全面的标准。最后，我们证明只有通过迁移学习和对非常深层进行微调才能获得模型自信心。
URL
https://arxiv.org/abs/2507.16779
