Search

Speech emotion recognition is a challenging topic and has many important applications in our real life, especially in terms of human-computer interaction. Traditional methods are based on the pipeline of pre-processing, feature extraction, dimensionality reduction and emotion classification. Previous studies have focussed on emotion recognition based on two different models: discrete model and continuous model. Both the speaker's age and gender affect the speech emotion recognition in the two models. Moreover, investigation results shown that the dimensional attributes of emotion such as arousal, valence and dominance are related to each other. Based on these observations, we propose a new attributes recognition model using Feature Nets, aims to improve the emotion recognition performance and generalisation capabilities. The method utilises the corpus to train the age and gender classification model, which will be transferred to the main model: a hierarchical deep learning model, using age and gender as the high level attributes of the emotion. The public databases EMO-DB and IEMOCAP have been conducted to evaluate the performance both in the classification task and regression task. Experiment results show that the proposed approach based on attributes transferring can improve the recognition accuracy, no matter transferring age or gender. [ABSTRACT FROM AUTHOR]

Although clinical outcomes of CLL have improved with the use of BCL-2 inhibitor, ABT-199, acquired resistance eventually occurs in many cases, which leads to CLL disease progression. Thus, understanding the mechanisms that mediate this relapse is important to design improved therapies. Herein, we report that cytokine IFN-γ, secreted by dysfunctional T cells, enhanced CLL cells resistance to ABT-199. IFN-γ stimulation significantly increased the expression of BCL-2, MCL-1 and BCL-xL. Blocking JAK1/2-STAT3 signaling pathway impaired the expression of these anti-apoptotic proteins after IFN-γ stimulation. The combination of ABT-199 with JAK1/2 inhibitor Ruxolitinib or STAT3 inhibitors Stattic and C188-9 increased malignant B cell death. In summary, we show that IFN-γ enhanced CLL cells resistance to ABT-199 at least in part by up-regulating BCL-2, MCL-1 and BCL-xL expression via JAK1/2-STAT3 pathway, and thus blocking this pathway with inhibitors increased ABT-199 efficiency to induce CLL cell apoptosis, suggesting a potential therapeutically relevant combination to overcome ABT-199 resistance. [ABSTRACT FROM AUTHOR]

With the capability of additive manufacturing, complex structures are easily fabricated to achieve various design purposes. In this work, a bi-material strip temperature sensor with complex periodic pattern design is purposed and investigated through both the analytical modeling and multi-physics finite element analysis. Three design patterns are considered: standard, E-shape and S-shape. In the standard solid strip design, the curvature of the bi-material strip under temperature variation is in a linear relationship with the coefficient of thermal expansion (CTE) difference, but in a reciprocal relationship with the strip thickness. The curvature of the bi-material strip depends on the Young's modulus ratio and layer thickness ratio of the two materials, but is independent of the magnitude of the materials' Young's modulus. Based on analytical derivation and numerical validation, the optimized design parameters can be provided. Compared to S-shape pattern design, E-shape pattern design can significantly increase the temperature sensitivity of the bi-material strip. An analytical prediction of the E-shape pattern's temperature sensitivity is introduced and discussed. This work proves the concept that new design space becomes available with the capability of additive manufacturing, and provides the general design guideline for a bi-material strip based temperature sensor with possible design patterns. [ABSTRACT FROM AUTHOR]