Estimation of a priori signal-to-noise ratio using neurograms for speech enhancement.

In statistical-based speech enhancement algorithms, the a priori signal-to-noise ratio (SNR) must be estimated to calculate the required spectral gain function. This paper proposes a method to improve this estimation using features derived from the neural responses of the auditory-nerve (AN) system. The neural responses, interpreted as a neurogram (NG), are simulated for noisy speech using a computational model of the AN system with a range of characteristic frequencies (CFs). Two machine learning algorithms were explored to train the estimation model based on NG features: support vector regression and a convolutional neural network. The proposed estimator was placed in a common speech enhancement system, and three conventional spectral gain functions were employed to estimate the enhanced signal. The proposed method was tested using the NOIZEUS database at different SNR levels, and various speech quality and intelligibility measures were employed for performance evaluation. The a priori SNR estimated from NG features achieved better quality and intelligibility scores than that of recent estimators, especially for highly distorted speech and low SNR values.

Measuring vocal difference in bird population pairs.

Over time, a bird population's acoustic and morphological features can diverge from the parent species. A quantitative measure of difference between two populations of species/subspecies is extremely useful to zoologists. Work in this paper takes a dialect difference system first developed for speech and refines it to automatically measure vocalisation difference between bird populations by extracting pitch contours. The pitch contours are transposed into pitch codes. A variety of codebook schemes are proposed to represent the contour structure, including a vector quantization approach. The measure, called Bird Vocalisation Difference, is applied to bird populations with calls that are considered very similar, very different, and between these two extremes. Initial results are very promising, with the behaviour of the metric consistent with accepted levels of similarity for the populations tested to date. The influence of data size on the measure is investigated by using reduced datasets. Results of species pair classification using Gaussian mixture models with Mel-frequency cepstral coefficients is also given as a baseline indicator of class confusability.

ViSQOLAudio: An objective audio quality metric for low bitrate codecs.

Streaming services seek to optimise their use of bandwidth across audio and visual channels to maximise the quality of experience for users. This letter evaluates whether objective quality metrics can predict the audio quality for music encoded at low bitrates by comparing objective predictions with results from listener tests. Three objective metrics were benchmarked: PEAQ, POLQA, and VISQOLAudio. The results demonstrate objective metrics designed for speech quality assessment have a strong potential for quality assessment of low bitrate audio codecs.

Deep Multi-Scale Feature Learning for Defocus Blur Estimation.

This paper presents an edge-based defocus blur estimation method from a single defocused image. We first distinguish edges that lie at depth discontinuities (called depth edges, for which the blur estimate is ambiguous) from edges that lie at approximately constant depth regions (called pattern edges, for which the blur estimate is well-defined). Then, we estimate the defocus blur amount at pattern edges only, and explore an interpolation scheme based on guided filters that prevents data propagation across the detected depth edges to obtain a dense blur map with well-defined object boundaries. Both tasks (edge classification and blur estimation) are performed by deep convolutional neural networks (CNNs) that share weights to learn meaningful local features from multi-scale patches centered at edge locations. Experiments on naturally defocused images show that the proposed method presents qualitative and quantitative results that outperform state-of-the-art (SOTA) methods, with a good compromise between running time and accuracy.

How does drinking water affect attention and memory? The effect of mouth rinsing and mouth drying on children's performance.

There is general consensus that drinking water facilitates certain cognitive processes. However, it is not yet known what mechanism underlies the effect of drinking on performance and these may be different for different cognitive processes. We sought to elucidate the mechanisms involved by establishing at what stage of the drinking process cognitive performance is influenced. We examined the effect of mouth rinsing and mouth drying on subjective thirst and mood, visual attention and short term memory in children. Data are reported from 24 children aged 9- to 10-years. Children's performance was assessed in three conditions - mouth drying, mouth rinsing and a control (no intervention). In each condition they were assessed twice - at baseline, before intervention, and 20 min later at test. Mouth rinsing improved visual attention performance, but not short term memory, mood or subjective thirst. The effects of mouth drying were more equivocal. The selective nature of the results is consistent with suggestions that different domains of cognition are influenced by different mechanisms.

Algorithms for the digital restoration of torn films.

This paper presents algorithms for the digital restoration of films damaged by tear. As well as causing local image data loss, a tear results in a noticeable relative shift in the frame between the regions at either side of the tear boundary. This paper describes a method for delineating the tear boundary and for correcting the displacement. This is achieved using a graph-cut segmentation framework that can be either automatic or interactive when automatic segmentation is not possible. Using temporal intensity differences to form the boundary conditions for the segmentation facilitates the robust division of the frame. The resulting segmentation map is used to calculate and correct the relative displacement using a global-motion estimation approach based on motion histograms. A high-quality restoration is obtained when a suitable missing-data treatment algorithm is used to recover any missing pixel intensities.

Simulated performance intensity functions.

Measuring speech intelligibility for different hearing aid fitting methods in a simulated environment would allow rapid prototyping and early design assessment. A simulated performance intensity function (SPIF) test methodology has been developed to allow experimentation using an auditory nerve model to predict listeners' phoneme recognition. The test discriminates between normal hearing and progressively degrading levels of sensorineural hearing loss. Auditory nerve discharge patterns, presented as neurograms, can be subjectively ranked by visual inspection. Here, subjective inspection is substituted with an automated ranking using a new image similarity metric that can quantify neurogram degradation in a consistent manner. This work reproduces the test results of a real human listener with moderate hearing loss, in unaided and aided scenarios, using a simulation. The simulated results correlate within comparable error margins to the real listener test performance intensity functions.