The effects of mood on emotion recognition and its relationship with the global vs local information processing styles
In this paper we consider the automatic emotions recognition problem, especially the case of digital audio signal processing. We consider and verify an approach in which the classification of a sound fragment is reduced to the problem of image recognition. The waveform and spectrogram are used as a visual representation of the image. The computational experiment was done based on Radvess open dataset including 8 different emotions: "neutral", "calm", "happy," "sad," "angry," "scared", "disgust", "surprised". The best accuracy result was 64%, which was produced by a combination of “|spectrogram + convolution neural network VGG-11”
Linguistic problems of individuals with agrammatic aphasia are not solely restricted to the grammatical domain: a considerable delay in lexical processing was also found in this clinical population (Prather et al., 1997). It was suggested that language processing abilities of aphasic individuals is predictable from their working memory (WM) capacities (Caspari et al., 1998; Friedmann & Gvion, 2003; Wright & Fergadiotos, 2012), however experimental evidence for that is still sparse.
The goal of the present study was to investigate the time course of lexical ambiguity resolution in healthy individuals and patients with agrammatism as a function of their WM span. We hypothesized that patients’ poorer than overall normal performance could at least partly be explained by their reduced WM capacities. Specifically, patients and healthy low WM span individuals were expected to demonstrate similar processing strategies.
Inefficient lexical processing has been found in both fluent and non-fluent aphasia, although different underlying mechanisms were proposed for those two clinical populations (Prather et al., 1997). Individuals with non-fluent aphasia were suggested to have delayed initial lexical activation, while problems of individuals with fluent aphasia concerned inhibition of irrelevant activation. The current study was designed to further tap into the time course of lexical processing in non-fluent and fluent aphasia. Specifically, lexical ambiguity resolution was investigated using an eye‐tracking‐while‐listening paradigm.
We investigated automatic Spatial–Numerical Association of Response Codes (SNARC) effect in auditory number processing. Two experiments continually measured spatial characteristics of ocular drift at central fixation during and after auditory number presentation. Consistent with the notion of a spatially oriented mental number line, we found spontaneous magnitude-dependent gaze adjustments, both with and without a concurrent saccadic task. This fixation adjustment (1) had a small-number/left-lateralized bias and (2) it was biphasic as it emerged for a short time around the point of lexical access and it received later robust representation around following number onset. This pattern suggests a two-step mechanism of sensorimotor mapping between numbers and space — a first-pass bottom-up activation followed by a top-down and more robust horizontal SNARC. Our results inform theories of number processing as well as simulation-based approaches to cognition by identifying the characteristics of anoculomotor resonance phenomenon.
In this paper we consider the automatic emotions recognition problem, especially the case of digital audio signal processing. We consider and verify an straight forward approach in which the classification of a sound fragment is reduced to the problem of image recognition. The waveform and spectrogram are used as a visual representation of the image. The computational experiment was done based on Radvess open dataset including 8 different emotions: “neutral”, “calm”, “happy,” “sad,” “angry,” “scared”, “disgust”, “surprised”. Our best accuracy result 71% was produced by combination “melspectrogram + convolution neural network VGG-16”.