Partially Supervised Speaker Clustering Hao Tang; Chu, S. M.; Hasegawa-Johnson, M. ...
IEEE transactions on pattern analysis and machine intelligence,
05/2012, Letnik:
34, Številka:
5
Journal Article
Recenzirano
Content-based multimedia indexing, retrieval, and processing as well as multimedia databases demand the structuring of the media content (image, audio, video, text, etc.), one significant goal being ...to associate the identity of the content to the individual segments of the signals. In this paper, we specifically address the problem of speaker clustering, the task of assigning every speech utterance in an audio stream to its speaker. We offer a complete treatment to the idea of partially supervised speaker clustering, which refers to the use of our prior knowledge of speakers in general to assist the unsupervised speaker clustering process. By means of an independent training data set, we encode the prior knowledge at the various stages of the speaker clustering pipeline via 1) learning a speaker-discriminative acoustic feature transformation, 2) learning a universal speaker prior model, and 3) learning a discriminative speaker subspace, or equivalently, a speaker-discriminative distance metric. We study the directional scattering property of the Gaussian mixture model (GMM) mean supervector representation of utterances in the high-dimensional space, and advocate exploiting this property by using the cosine distance metric instead of the euclidean distance metric for speaker clustering in the GMM mean supervector space. We propose to perform discriminant analysis based on the cosine distance metric, which leads to a novel distance metric learning algorithm-linear spherical discriminant analysis (LSDA). We show that the proposed LSDA formulation can be systematically solved within the elegant graph embedding general dimensionality reduction framework. Our speaker clustering experiments on the GALE database clearly indicate that 1) our speaker clustering methods based on the GMM mean supervector representation and vector-based distance metrics outperform traditional speaker clustering methods based on the "bag of acoustic features" representation and statistical model-based distance metrics, 2) our advocated use of the cosine distance metric yields consistent increases in the speaker clustering performance as compared to the commonly used euclidean distance metric, 3) our partially supervised speaker clustering concept and strategies significantly improve the speaker clustering performance over the baselines, and 4) our proposed LSDA algorithm further leads to state-of-the-art speaker clustering performance.
•Heterogeneous catalysis has emerged as a pivotal way for the plastic upcycling.•We summarized support effects on catalytic recycling of waste plastics.•We offered an outlook for the heterogeneous ...catalytic plastic recycling.
Heterogeneous supported metal catalysts have played a pivotal role in modern industrial initiatives dedicated to the chemical upcycling of end-of-life plastics. Leveraging the multifaceted effects of supports to customize the structural features of catalysts constitutes a widely embraced and effective strategy for enhancing catalytic performance. Nonetheless, the nascent field of plastics chemical upcycling lags in progress, primarily attributed to the intricate and versatile nature of these support effects. In this review, we provide a comprehensive overview of the current state of research into diverse support effects employed in the context of plastic upcycling. We delve deeply into the structure-performance relationship, scrutinizing key aspects such as metal-support interactions (MSI), promoter effects, support-polymer interactions, and bifunctional catalysis. Finally, we offer a forward-looking discourse, presenting an outlook for the advancement of heterogeneous catalysts built upon the amalgamation of the identified support effects and exploring potential auxiliary functions of these support materials. This discussion serves to provide essential guidance for the design of multifunctional catalysts with a focus on enhancing the process of plastic upcycling.
The conversion of CO2 into fuels and feedstock chemicals via photothermal catalysis holds promise for efficient solar energy utilization to tackle the global energy shortage and climate change. ...Despite recent advances, it is of emerging interest to explore promising materials with excellent photothermal properties to boost the performance of photothermal CO2 catalysis. Here, we report the discovery of MXene materials as superior photothermal supports for metal nanoparticles. As a proof-of-concept study, we demonstrate that Nb2C and Ti3C2, two typical MXene materials, can enhance the photothermal effect and thus boost the photothermal catalytic activity of Ni nanoparticles. A record CO2 conversion rate of 8.50 mol·gNi –1·h–1 is achieved for Nb2C-nanosheet-supported Ni nanoparticles under intense illumination. Our study bridges the gap between photothermal MXene materials and photothermal CO2 catalysis toward more efficient solar-to-chemical energy conversions and stimulates the interest in MXene-supported metal nanoparticles for other heterogeneous catalytic reactions, particularly driven by sunlight.
Background
Circular RNAs (circRNAs) have been identified to be involved in onset and progression of multiple malignant tumors. The present study aimed to systematically evaluate the diagnostic values ...of circRNAs in breast cancer.
Methods
The PubMed, Web of Science, Embase, CNKI, and Wanfang online databases were searched for the relevant studies before December 31, 2020. Statistical analysis of the diagnostic tests was performed based on STATA 16.0, Meta‐DiSc 1.4, and RevMan 5.3 software. The threshold effect and publication bias were measured by the Spearman correlation and Deeks’ funnel plot asymmetry test, respectively.
Results
Twenty‐one studies from 13 articles were included in this meta‐analysis. The pooled sensitivity and specificity were 0.77 and 0.71, respectively. The pooled positive likelihood ratio (PLR), negative likelihood ratio (NLR), and overall diagnostic odds ratio (DOR) were 2.6, 0.33, and 8, respectively. Furthermore, the area under the summary receiver operator characteristic curve was 0.80. In addition, down‐regulated circRNAs achieved a diagnostic performance higher than up‐regulated circRNAs, with area under curve (AUC) values of 0.81 and 0.74, respectively. Studies based on tissue samples presented better diagnostic accuracy than those based on plasma samples, with AUC values of 0.80 and 0.67. In addition, two circRNAs, including circ_0001073 and circTADA2A‐E5/E6, showed higher diagnostic values, with AUC value of 0.990 and 0.937, respectively. According to the results of meta‐regression, the case size (p<0.05) might be the source of the heterogeneity.
Conclusion
CircRNAs exhibited a high diagnostic value for breast cancer and may function as potential diagnostic biomarkers for breast cancer.
Thirteen studies comprised of 21 circRNAs were included in this meta‐analysis. Down‐regulated circRNAs achieved a diagnostic performance higher than up‐regulated circRNAs. Tissue‐based circRNA analysis presented better diagnostic accuracy than plasma‐based analysis.
Display omitted
•A strategy for regenerating sintered Pt catalyst at mild temperature is proposed.•The strategy is realized with the assistance of H2O2 and HCl.•This strategy has been implemented on ...Pt-Ga2O3/S40 and Pt-Sn/Al2O3 for PDH.•The strategy is effective for sintered Pd-, Au-, Ir-, Rh-, and Ru at mild temperature.
Pt has been extensively used as active site for propane dehydrogenation (PDH) to propene, a crucial important raw material in chemical industry. Nevertheless, industrial Pt-based catalyst for PDH suffers from sintering at high temperature, which requires a harsh regeneration process in the presence of purified air and hydrochloric acid (HCl) at high temperature (500 °C). Herein, we propose a robust strategy for regenerating sintered Pt-based catalyst with hydrogen peroxide (H2O2) and HCl at mild temperature (70 °C), where the oxidation with H2O2 and the chlorination with HCl can synergistically weaken Pt − Pt bond to form soluble species, realizing the redispersion of sintered Pt. The current strategy has been successfully implemented on Pt-Ga2O3/S40 and industrial Pt-Sn/Al2O3 for PDH, where the optimal propane conversion reaches > 72 % at a propene selectivity of ∼ 96 % at 600 °C, approaching the thermodynamic equilibrium. Impressively, the universality of the current strategy has been validated by regenerating sintered Pd-, Au-, Ir-, Rh-, and Ru at mild temperature. We believe this work provides a robust and industrially viable strategy for regenerating noble metal-based catalysts, which will capture extensive and immediate attention in chemical industry.
The reverse water–gas shift (RWGS) reaction is a critical step for adjusting the CO/H2 ratio in crude syngas via CO2 hydrogenation in the chemical industry. Generally, the conditioning of syngas ...competitively occurs in two different reaction pathways at a moderate temperature range (350–450 °C), such as RWGS (CO2 + H2 → CO + H2O) and methanation (CO2 + 4H2 → CH4 + 2H2O). In this study, we demonstrate that the interfacial synergy of Ni/Ga2O3 can boost selective CO2 hydrogenation to CO (CO selectivity >95%) in the temperature range of 350–450 °C. Detailed investigations indicate that the interfacial synergy of the Ni/Ga2O3 interface significantly influences the adsorption of H2 and CO2 and thus leads to the formation of different intermediates and products. In particular, *HCOO is preferentially formed on the Ni surface, which is further hydrogenated into CH4 and H2O. By contrast, the Ni/Ga2O3 interface favors the formation of CO with the assistance of H2. This study not only provides a highly selective catalyst for the RWGS reaction during CO2 hydrogenation but also promotes research on surface modification of catalysts for enhanced activity and selectivity.
Revealing the relationships between geometric structure and electrochemical stability of Pt-based nanostructures is significant to catalyst development for the ethanol oxidation reaction (EOR) in an ...acidic media. Herein, we proposed the controllable construction of Pt x Bi y nanostructures with a continuous structural evolution by a simple colloidal synthesis. As the Bi feeding decreased, the inhibition effect of Bi3+ on the epitaxial growth of Pt nanoparticles (NPs) was diminished, and Pt x Bi y nanostructures transformed from nanoparticle aggregates to nanodendrites, which induced the generation of PtBi–Bi2O3 heterogeneous nanostructures by the electrochemical reactivation process. Benefiting from the modification of Bi species and nanodendrites, the optimal Pt2Bi1 catalysts showed superior structural stability and catalytic durability with a high mass activity of 482.6 mA mgPt –1, which was 2.2 times that of commercial Pt/C. This work not only prepared a series of Pt-based heterogeneous nanostructures, but also provided an effective approach for high-performance EOR based on the structure–performance relationship.
Single atom catalysts (SACs) have recently attracted great attention in heterogeneous catalysis and have been regarded as ideal models for investigating the strong interaction between metal and ...support. Despite the huge progress over the past decade, the deep understanding on the structure-performance correlation of SACs at a single atom level still remains to be a great challenge. In this study, we demonstrate that the variation in the coordination number of the Pt single atom can significantly promote the propylene selectivity during propyne semihydrogenation (PSH) for the first time. Specifically, the propylene selectivity greatly increases from 65.4% to 94.1% as the coordination number of Pt–O increases from ∼3.4 to ∼5, whereas the variation in the coordination number of Pt–O slightly influences the turnover frequency values of SACs. We anticipate that the present work may deepen the understanding on the structure-performance of SACs and also promote the fundamental research in single atom catalysis.
Abstract
Liquid hydridopolycarbosilanes (LHPCS) with vinyl or allyl groups have attracted significant research interest due to their low curing temperature and high ceramic yield. In this study, ...allyl‐terminated hydridopolycarbosilane (AHPCS) and vinyl‐terminated hydridopolycarbosilane (VHPCS) were prepared by the Grignard coupling reaction and characterized by NMR and Fourier infrared spectroscopy (FTIR). And the curing reactions of AHPCS and VHPCS with or without initiators (dicumyl peroxide DCP, tert‐Butyl peroxybenzoate TBPB) or Karstedt catalyst were investigated using differential scanning calorimetry and in‐situ FTIR. The main curing reactions of VHPCS and AHPCS are the double bond free radical polymerization and hydrosilylation reaction, respectively. The addition of initiators and Karstedt catalyst could promote the corresponding curing reaction and increase the weight residual at 1000°C. The thermal properties of the cured resins were investigated by thermogravimetric analysis. The addition of initiators and catalyst could significantly increase the weight residual of VHPCS and AHPCS, respectively. The addition of the Karstedt catalyst increased the weight residual of AHPCS from 75.6% to 83.1%. In addition, the addition of TBPB and DCP increased the weight residual of VHPCS from 74.2% to 81.2% and 80.2%, respectively.