In skeleton-based abnormal gait recognition, using original skeleton data decreases the recognition performance because they contain noise and irrelevant information. Instead of feeding original skeletal gait data to a recognition model, features extracted from the skeleton data are normally used. However, existing feature extraction methods might include laborious processes and it is hard for them to minimize the irrelevant information while preserving the important information. To solve this problem, an automatic feature extraction method using a recurrent neural network (RNN)-based Autoencoder (AE) is proposed in this paper. We extracted features from skeletal gait data by using two RNN AEs: a long shortterm memory (LSTM)-based AE (LSTM AE) and a gated recurrent unit (GRU)-based AE (GRU AE). The features of the RNN AEs are compared to the original skeleton data and other existing features. We evaluated the features by feeding them to various discriminative models (DMs) and comparing the recognition performances. The features extracted by using the RNN AEs are more easily recognized and robust than the original skeleton data and other existing features. In particular, the LSTM AE shows a better performance than the GRU AE. Compared to single DMs fed with the original skeleton directly, hybrid models where the features of the RNN AEs are fed to DMs show a higher recognition accuracy with fewer training epochs and learning parameters. Therefore, the proposed automatic feature extraction method improves the performance of skeleton-based abnormal gait recognition by reducing laborious processes and increasing the recognition accuracy effectively.

Herpes zoster (HZ) is a common cutaneous disease affecting one out of five people; hence, early diagnosis of HZ is crucial as it can progress to chronic pain syndrome if antiviral treatment is not provided within 72 hr. Mobile diagnosis of HZ with the assistance of artificial intelligence can prevent neuropathic pain while reducing clinicians’ fatigue and diagnosis cost. However, the clinical images captured from daily mobile devices likely contain visual corruptions, such as motion blur and noise, which can easily mislead the automated system. Hence, this paper aims to train a robust and mobile deep neural network (DNN) that can distinguish HZ from other skin diseases using user-submitted images. To enhance robustness while retaining low computational cost, we propose a knowledge distillation from ensemble via curriculum training (KDE-CT) wherein a student network learns from a stronger teacher network progressively. We established skin diseases dataset for HZ diagnosis and evaluated the robustness against 75 types of corruption. A total of 13 different DNNs was evaluated on both clean and corrupted images. The experiment result shows that the proposed KDE-CT significantly improves corruption robustness when compared with other methods. Our trained MobileNetV3-Small achieved more robust performance (93.5% overall accuracy, 67.6 mean corruption error) than the DNN ensemble with smaller computation (549x fewer multiply-and-accumulate operations), which makes it suitable for mobile skin lesion analysis.

A simple, novel, rapid, and non-destructive spectroscopic method that employs the deep spectral network for beef-freshness classification was developed. The deep-learning-based model classified beef freshness by learning myoglobin information and reflectance spectra over different freshness states. The reflectance spectra (480-920 nm) were measured from 78 beef samples for 17 days, and the datasets were sorted into three freshness classes based on their pH values. Myoglobin information showed statistically significant differences depending on the freshness; consequently, it was utilized as a crucial parameter for classification. The model exhibited improved performance when the reflectance spectra were combined with the myoglobin information. The accuracy of the proposed model improved to 91.9%, whereas that of the single-spectra model was 83.6%. Further, a high value for the area under the receiver operating characteristic curve (0.958) was recorded. This study provides a basis for future studies on the investigation of myoglobin information associated with meat freshness.

Dental implantation is a surgical procedure in oral and maxillofacial surgery. Detecting missing tooth regions is essential for planning dental implant placement. This study proposes an automated method that detects regions of missing teeth in panoramic radiographic images. Tooth instance segmentation is required to accurately detect a missing tooth region in panoramic radiographic images containing obstacles, such as dental appliances or restoration. Therefore, we constructed a dataset that contains 455 panoramic radiographic images and annotations for tooth instance segmentation and missing tooth region detection. First, the segmentation model segments teeth into the panoramic radiographic image and generates teeth masks. Second, a detection model uses the teeth masks as input to predict regions of missing teeth. Finally, the detection model identifies the position and number of missing teeth in the panoramic radiographic image. We achieved 92.14% mean Average Precision (mAP) for tooth instance segmentation and 59.09% mAP for missing tooth regions detection. As a result, this method assists diagnosis by clinicians to detect missing teeth regions for implant placement.

Extraction of mandibular third molars is a common procedure in oral and maxillofacial surgery. There are studies that simultaneously predict the extraction difficulty of mandibular third molar and the complications that may occur. Thus, we propose a method of automatically detecting mandibular third molars in the panoramic radiographic images and predicting the extraction difficulty and likelihood of inferior alveolar nerve (IAN) injury. Our dataset consists of 4903 panoramic radiographic images acquired from various dental hospitals. Seven dentists annotated detection and classification labels. The detection model determines the mandibular third molar in the panoramic radiographic image. The region of interest (ROI) includes the detected mandibular third molar, adjacent teeth, and IAN, which is cropped in the panoramic radiographic image. The classification models use ROI as input to predict the extraction difficulty and likelihood of IAN injury. The achieved detection performance was 99.0% mAP over the intersection of union (IOU) 0.5. In addition, we achieved an 83.5% accuracy for the prediction of extraction difficulty and an 81.1% accuracy for the prediction of the likelihood of IAN injury. We demonstrated that a deep learning method can support the diagnosis for extracting the mandibular third molar.