Combining missing-feature theory, speech enhancement, and speaker-dependent/-independent modeling for speech separation

This paper considers the separation and recognition of overlapped speech sentences assuming single-channel observation. A system based on a combination of several different techniques is proposed. The system uses a missing-feature approach for improving crosstalk/noise robustness, a Wiener filter for speech enhancement, hidden Markov models for speech reconstruction, and speaker-dependent/-independent modeling for speaker and speech recognition. We develop the system on the Speech Separation Challenge database, involving a task of separating and recognizing two mixing sentences without assuming advanced knowledge about the identity of the speakers nor about the signal-to-noise ratio. The paper is an extended version of a previous conference paper submitted for the challenge.

A latent feature analysis of the neural representation of conceptual knowledge

Bayesian probabilistic analysis offers a new approach to characterize semantic representations by inferring the most likely feature structure directly from the patterns of brain activity. In this study, infinite latent feature models [1] are used to recover the semantic features that give rise to the brain activation vectors when people think about properties associated with 60 concrete concepts. The semantic features recovered by ILFM are consistent with the human ratings of the shelter, manipulation, and eating factors that were recovered by a previous factor analysis. Furthermore, different areas of the brain encode different perceptual and conceptual features. This neurally-inspired semantic representation is consistent with some existing conjectures regarding the role of different brain areas in processing different semantic and perceptual properties. © 2012 Springer-Verlag.

Who was Andrea? Writing Oneself as a Feminist Icon

This article explores the construction of Andrea Dworkin as a public persona, or a ‘feminist icon’, revered by some and demonized by others. It argues that in both her fiction and non-fiction, Dworkin engaged in a process of writing herself as an exceptional woman, a ‘feminist militant’ as she describes herself in the subheading of her 2002 memoir, Heartbreak. The article illustrates Dworkin’s autobiographical logic of exceptionalism by comparing the story told in Heartbreak to the story of Dworkin’s major novel, Mercy, which features a heroine, Andrea, who shares Dworkin’s name and significant biographical details. While Dworkin has insisted that Mercy is not an autobiographical novel, the author undertakes a reading here of Mercy as the story of Dworkin if she had not become the feminist icon of her own and others’ construction. In Mercy, Andrea unsuccessfully attempts to escape the silent, victimized status that Dworkin has insistently argued is imposed upon women. In her repeated victimization, Andrea functions for Dworkin as an ‘everywoman’ who both embodies Dworkin’s world-view and highlights how Dworkin’s own biography exists in tension with some of her central assumptions about women, gender and contemporary society.

Water-mediated electrochemical nano-writing on thin ceria films

Bias dependent mechanisms of irreversible cathodic and anodic processes on a pure CeO2 film are studied using modified atomic force microscopy (AFM). For a moderate positive bias applied to the AFM tip an irreversible electrochemical reduction reaction is found, associated with significant local volume expansion. By changing the experimental conditions we are able to deduce the possible role of water in this process. Simultaneous detection of tip height and current allows the onset of conductivity and the electrochemical charge transfer process to be separated, further elucidating the reaction mechanism. The standard anodic/cathodic behavior is recovered in the high bias regime, where a sizable transport current flows between the tip and the film. These studies give insight into the mechanisms of the tip-induced electrochemical reactions as mediated by electronic currents, and into the role of water in these processes, as well as providing a different approach for electrochemical nano-writing.