Audio Context Research Articles

Proffering a discourse in different communicative contexts

Communicative contexts may affect how the speaker proffers a discourse. In particular, we assumed that uni-directional (as compared with bi-directional) and audio (as compared with audio-visual) contexts induce the speaker to elaborate and then use an articulated mental model of the discourse because they do not allow the exploitation of all the communicative means. Uni-directional contexts do not allow recovery of communicative failures, and audio contexts do not allow access to extralinguistic communication. The results of an experiment involving 84 adult participants confirmed the predictions deriving from these assumptions: linguistic indices of the exploitation of an articulated mental model of the discourse are greater in uni-directional and audio contexts as compared with bi-directional and audio-visual contexts, respectively.

Initialization, distances and local minima in audio applications of the non‐negative matrix factorization

The use of the non‐negative matrix factorization (NMF) as a decomposition technique has dramatically grown in various signal processing applications over the last years. Its computation, based on the iterative minimization of a cost function, relies on several choices, among which the distance involved in the cost function itself but also the initialization of the algorithm. These choices are all the more crucial as the usual algorithms, iterative, only ensure convergence to a local minimum. In this work, we compare three typical distances in the NMF problem (namely Euclidian, Kullback‐Leibler divergence and Itakura‐Saito distance) and their combination with different initializations, in an audio context: decomposition of the time‐frequency representation for the transcription of polyphonic music. Both the existence of global and local minima, and the efficiency of transcription are examined. Moreover, NMF update rules can be formulated in a unified framework for the three aforementioned cost functions. This formulation allows figuring out new NMF algorithms which could address the local minima question.

Amen to That

Amen to That

The Impact of Interstimulus Interval and Background Silence on Recall

Three experiments investigate how the interval between target items of information in an audio context influences recall under background conditions of music and silence. In experiment 1, when a goal-directed learning task encouraged allocation of resources to the message, recall increased linearly for both background conditions as the interval increased. However, under the incidental learning task used in experiment 2, increasing the interval from two to three seconds had a deleterious impact on recall, but only in the background silence condition. Experiment 3 suggested that this was due to a surplus of cognitive resources in the background silence condition producing interfering extracommunication thought. Results also support the theory presented that background music borrows resources from the processing of message information in an incidental learning task. Although music impairs processing of brand information and reduces recall at shorter intervals, relative to silence, it increases the interval preceding the onset of excess resources, delaying interfering thought.

Capturing, structuring, and representing ubiquitous audio

Although talking is an integral part of collaboration, there has been little computer support for acquiring and accessing the contents of conversations. Our approach has focused on ubiquitous audio , or the unobtrusive capture of speech interactions in everyday work environments. Speech recognition technology cannot yet transcribe fluent conversational speech, so the words themselves are not available for organizing the captured interactions. Instead, the structure of an interaction is derived from acoustical information inherent in the stored speech and augmented by user interaction during or after capture. This article describes applications for capturing and structuring audio from office discussions and telephone calls, and mechanisms for later retrieval of these stored interactions. An important aspect of retrieval is choosing an appropriate visual representation, and this article describes the evolution of a family of representations across a range of applications. Finally, this work is placed within the broader context of desktop audio, mobile audio applications, and social implications.