Scanning transmission electron microscopy (STEM) can perform electron imaging of material properties at the atomic picometer level and interpret the atomic structure using the obtained images. However, obtaining high-quality atomic-scale STEM images requires advanced STEM equipment and skilled operators. Various environmental factors can introduce unpredictable non-uniform noise during the STEM imaging process, thereby significantly affecting image quality and consequently influencing the results of atomic structure analysis. The prediction model based on deep neural networks can reduce the impact of noise through denoising or data fitting, but there exists a problem of overfitting. This paper introduces materials structure conditions as priors in the deep neural network model and designs a method for atomic structure segmentation of high-noise STEM images based on materials structural priors. In this method, the materials structural priors are modelled as the attention (including self-attention and cross-attention) of the segmentation network and are calculated, which not only enables the segmentation network to adaptively focus on the key regions of the image but also to adaptively focus on the control information from the structural coordinate vector modalities. In the simulation test set, as compared with AtomAI Segmentor method, the proposed method improves the chamfer distance, Jaccard and F1 metrics by 175%, 49.7% and 42.7%, respectively; as compared with the early multi-scale method proposed by the research group, it improves the chamfer distance, Jaccard and F1 metrics by 167%, 28% and 23.9%, respectively. In the laboratory sample test set, as compared with AtomAI Segmentor method, the proposed method improves the chamfer distance, Jaccard and F1 metrics by 63%, 9.3% and 7.4%, respectively; as compared with the early multi-scale method proposed by the research group, it improves the chamfer distance by 12.8%, and the Jaccard and F1 metrics remain largely unchanged. The introduction of materials structural priors enables the segmentation network model to more accurately segment the atomic structure in high-noise STEM images and predict the secondary structure information that is affected by noise or top-level occlusion.

Automatic detection of epilepsy based on electroencephalogram (EEG) signals is greatly helpful for clinical diagnosis and treatment of epilepsy. Most epilepsy detection algorithms ignore the temporal relation of EEG signals, therefore, this paper proposed an epilepsy EEG signal classification method based on multi-band path signature features. Firstly, EEG signals were decomposed into five frequency bands. Secondly, features were extracted using the path signature (PS) algorithm. Thirdly, features were fused after local principal component analysis (LPCA) removed the feature’s correlation. Finally, an ensemble classifier was used to predict epilepsy. Since the path signature can dig into the correlation of EEG signals, combined with local principal component analysis, the method proposed in the paper can obtain more discriminative epilepsy classification features. The comparative experiments of 10-fold cross-validation were conducted to validate this method on two datasets, i. e., the private dataset from a local hospital with more than 2000 seconds of segments and the CHB-MIT epilepsy dataset. The results show that the average classification accuracy of the method reached 97.25% on the private dataset, which is higher than those of the classical EMD (Empirical Mode Decomposition) method and the up-todate LSTM (Long Short-term Memory Network) + CNN (Convolutional Neural Network) method by 3.44 and 1.35 percentage points respectively. Moreover, the proposed method can achieve an average classification accuracy of 98.11% on the CHB-MIT dataset, which is higher than those of the classical EMD method and the up-to-date LSTM+CNN method by 5.20 and 2.64 percentage points respectively, and this method achieves the best classification accuracy than other methods on both datasets.