Multi-Input Speech Emotion Recognition Model Using Mel Spectrogram and GeMAPS

The existing research on emotion recognition commonly uses mel spectrogram (MelSpec) and Geneva minimalistic acoustic parameter set (GeMAPS) as acoustic parameters to learn the audio features. MelSpec can represent the time-series variations of each frequency but cannot manage multiple types of audio features. On the other hand, GeMAPS can handle multiple audio features but fails to provide information on their time-series variations. Thus, this study proposes a speech emotion recognition model based on a multi-input deep neural network that simultaneously learns these two audio features. The proposed model comprises three parts, specifically, for learning MelSpec in image format, learning GeMAPS in vector format, and integrating them to predict the emotion. Additionally, a focal loss function is introduced to address the imbalanced data problem among the emotion classes. The results of the recognition experiments demonstrate weighted and unweighted accuracies of 0.6657 and 0.6149, respectively, which are higher than or comparable to those of the existing state-of-the-art methods. Overall, the proposed model significantly improves the recognition accuracy of the emotion "happiness", which has been difficult to identify in previous studies owing to limited data. Therefore, the proposed model can effectively recognize emotions from speech and can be applied for practical purposes with future development.

Human Hepatocellular Carcinoma Classification from H&E Stained Histopathology Images with 3D Convolutional Neural Networks and Focal Loss Function

This paper proposes a new Hepatocellular Carcinoma (HCC) classification method utilizing a hyperspectral imaging system (HSI) integrated with a light microscope. Using our custom imaging system, we have captured 270 bands of hyperspectral images of healthy and cancer tissue samples with HCC diagnosis from a liver microarray slide. Convolutional Neural Networks with 3D convolutions (3D-CNN) have been used to build an accurate classification model. With the help of 3D convolutions, spectral and spatial features within the hyperspectral cube are incorporated to train a strong classifier. Unlike 2D convolutions, 3D convolutions take the spectral dimension into account while automatically collecting distinctive features during the CNN training stage. As a result, we have avoided manual feature engineering on hyperspectral data and proposed a compact method for HSI medical applications. Moreover, the focal loss function, utilized as a CNN cost function, enables our model to tackle the class imbalance problem residing in the dataset effectively. The focal loss function emphasizes the hard examples to learn and prevents overfitting due to the lack of inter-class balancing. Our empirical results demonstrate the superiority of hyperspectral data over RGB data for liver cancer tissue classification. We have observed that increased spectral dimension results in higher classification accuracy. Both spectral and spatial features are essential in training an accurate learner for cancer tissue classification.

Optical Fiber Vibration Signal Identification Method Based on Improved YOLOv4

In the traditional peripheral-security-early-warning system, the endpoint detection and pattern recognition of the signals generated by the distributed optical fiber vibration sensors is completed step-by-step and in an orderly manner. The method by which these two processes may be placed end-to-end in a network model and processed simultaneously to improve work efficiency has increasingly become the focus of research. In this paper, the target detection algorithm combines the endpoint-detection and pattern-recognition processes of the vibration signal, which can not only quickly locate the start and end vibration positions of the signal but also accurately identify a certain type of signal. You Only Look Once v4 (YOLOv4) is one of the most advanced target detection algorithms, achieving the optimal balance of speed and accuracy. To reduce the complexity of the YOLOv4 model and solve the dataset's unbalanced sample classification problem, we use a deep separable convolution (DSC) network and a focal loss function to improve the YOLOv4 model. In this paper, the five kinds of signals collected in real-time are visualized as two different datasets in oscillograph and time-frequency diagrams as detection objects. According to the experimental results, we obtained 98.50% and 93.48% mean Average Precision (mAP) and 84.8 and 69.9 frames per second (FPS), respectively, which are improved compared to YOLOv4. Comparing the improved algorithm with other optical fiber vibration signal recognition algorithms, the mAP and FPS values were improved, and the detection speed was about 20 times faster than that of other algorithms. The improved algorithm in this paper can quickly and accurately identify the vibration signal of external intrusion, reduce the false-alarm rate of the early-warning system, and improve the real-time detection rate of the system while ensuring high recognition accuracy.

Automatic detection of calcium phosphate deposit plugs at the terminal ends of kidney tubules

Kidney stones are a common urologic condition with a high amount of recurrence. Recurrence depends on a multitude of factors the incidence of precursors to kidney stones, plugs, and plaques. One method of characterising the stone precursors is endoscopic assessment, though it is manual and time-consuming. Deep learning has become a popular technique for semantic segmentation because of the high accuracy that has been demonstrated. The present Letter examined the efficacy of deep learning to segment the renal papilla, plaque, and plugs. A U-Net model with ResNet-34 encoder was tested; the Letter examined dropout (to avoid overtraining) and two different loss functions (to address the class imbalance problem. The models were then trained in 1666 images and tested on 185 images. The Jaccard-cross-entropy loss function was more effective than the focal loss function. The model with the dropout rate 0.4 was found to be more effective due to its generalisability. The model was largely successful at delineating the papilla. The model was able to correctly detect the plaques and plugs; however, small plaques were challenging. Deep learning was found to be applicable for segmentation of an endoscopic image for the papilla, plaque, and plug, with room for improvement.