Confusion Patterns Research Articles

Asynchronous Vowel-Pair Identification Across the Adult Life Span for Monaural and Dichotic Presentations

Temporal order abilities decrease with age. Declining temporal processing abilities may influence the identification of rapid vowel sequences. Identification patterns for asynchronous vowel pairs were explored across the life span. Young, middle-aged, and older listeners completed temporal order tasks for pairs of 70-ms and 40-ms vowel stimuli. For a given vowel duration, naturally spoken vowels were equated for duration, intensity, and fundamental frequency. Listeners completed monaural and dichotic temporal order tasks that involved identifying the vowel pair in the correct order. The stimulus onset asynchrony that yielded 50% accuracy for identifying the vowel pair in the correct order was used to equate performance among listeners. Vowel identification response patterns were determined at this stimulus onset asynchrony threshold. Vowel identification patterns were largely consistent across age groups. Older listeners were influenced by the order of certain vowel pairs. Not all vowel pairs were identified equally well. Vowel dominance patterns were also observed, with /a/ being identified most accurately for the vowel pairs tested. Formant dynamics explained, in part, identification and confusion patterns. Vowel identification accuracy patterns were reasonably similar across the life span, regardless of presentation mode, vowel duration, or effect of considerable stimulus exposure. Large effects of vowel order were observed, particularly for older listeners.

Color vision in an elderly patient with protanopic genotype and successfully treated unilateral age-related macular degeneration

We investigated differences in color discrimination between the fellow eye and the affected eye successfully treated for unilateral age-related macular degeneration (AMD) in a 69-year-old male patient with protanopia. His best-corrected visual acuity (BCVA) was 1.2 in the right eye (RE) and 0.2 in the left eye (LE). Fundus and angiographic findings showed classic choroidal neovascularization (CNV) secondary to AMD in the LE. BCVA of the LE improved to 0.4, and CNV resolved by 15months after initiating combined anti-vascular endothelial growth factor and photodynamic therapies. After CNV closure, the Farnsworth dichotomous was performed, showing confusion patterns of the protan axis in either eye. The Farnsworth-Munsell 100-hue test showed a total error score of 520 in the LE, much higher than the score of 348 in the RE. Complete genotypes of the long-wavelength-sensitive (L-) cone and middle-wavelength-sensitive (M-) cone opsin genes were determined by polymerase chain reaction, revealing that the patient had a single 5' L-M 3' hybrid gene (encoding an M-cone opsin), with this genotype responsible for protanopia (the L-cone opsin gene was non-functional), instead of the L-cone and M-cone opsin gene arrays. Poorer color vision discrimination in the LE than the RE remained present despite closure of CNV. The presence and type of congenital color vision defect can be confirmed using molecular genetic testing even if complications of acquired retinal diseases such as AMD are identified.

What causes the confusion patterns in speech perception?

It is widely believed that speech contains redundant cues which make it robust in noise. We have discovered that natural speech often contains conflicting cues that cause significant confusion patterns in speech perception. To demonstrate the impact of conflicting cues on the identification of consonant sounds, we selected four plosives /ta, ka, da, ga/ and four fricatives /si, Si, zi, Zi/ that show significant confusion with other sounds from a female high-performance talker in the LDC database and derived a set of test stimuli by removing the dominant cue and manipulating the conflicting cue. It was found that the consonants always morph to their competing sounds, once the dominant cue is unavailable. The confusion can be strengthened by enhancing the conflicting cue. These results demonstrate that the identification of consonant is dependent on a careful balancing between the dominant cue and the conflicting cues. Such prior knowledge about perceptual cues can be used to manipulate the perception of consonants, as demonstrated by the selected examples of nonsense syllables, meaningful words, and sentence.

Perceptual deficits in Thai sensorineural hearing loss patients

This study explored differences in CVV perception in two groups of Thai listeners: with normal hearing and with sensorineural hearing loss (with/without hearing aids). All participants chose one response in each of 210 Thai stimulus rhyming pairs, e.g., /taa/-/naa/. The rhyming monosyllabic words share an /aa/ vowel and mid tone, but differ in their initial phonemes (symmetrically distributed across 21 phonemes). While all stimuli for the normal hearing group were embedded in 4 signal-to-noise ratio levels, clean stimuli were presented to the patients. Comparisons of confusion patterns and perceptual distance were made. In both groups, /r/ is the most confusable phoneme, while /w/ is among the least. Perceptual representations of initial phonemes show five individual clusters: glide, glottal constriction, nasality, aspirated obstruent, and a combination of liquid and unaspirated obstruent. Patients' perceptual difficulty could be attributed to the nasality grouping, which is normally well separated, shifting closer to the glottal constrictions and aspirated obstruents. Hearing aids seem to improve perception of all phonemes by 10%, with /kh/ and /h/ showing the highest improvement rate, and /d/ the lowest. The instruments are beneficial in moving the nasality cluster further away from the nearby groupings.

Orientation Information in Encoding Facial Expressions

Abstract Previous research showed that we use different regions of a face to categorize different facial expressions, e.g. mouth region for identifying happy faces; eyebrows, eyes and upper part of nose for identifying angry faces. These findings imply that the spatial information along or close to the horizontal orientation might be more useful than others for facial expression recognition. In this study, we examined how the performance for recognizing facial expression depends on the spatial information along different orientations, and whether the pixel-level differences in the face images could account for subjects’ performance. Four facial expressions—angry, fearful, happy and sad—were tested. An orientation filter (bandwidth = 23°) was applied to restrict information within the face images, with the center of the filter ranged from 0° (horizontal) to 150° in steps of 30°. Accuracy for recognizing facial expression was measured for an unfiltered and the six filtered conditions. For all four facial expressions, recognition performance (normalized d′) was virtually identical for filter orientations of −30°, horizontal and 30°, and declined systematically as the filter orientation approached vertical. The information contained in mouth and eye regions is a significant predictor for subject’s response (based on the confusion patterns). We conclude that young adults with normal vision categorizes facial expression most effectively based on the spatial information around the horizontal orientation which captures primary changes of facial features across expressions. Across all spatial orientations, the information contained in mouth and eye regions contributes significantly to facial expression categorization.

F065 Improved speech perception in noise for cochlear implantees using binaural signal processing

Noise-band vocoders are often used to simulate the signal processing algorithms used in cochlear implants (CIs), producing acoustic stimuli that may be presented to normal hearing (NH) subjects. Such evaluations may obviate the heterogeneity of CI user populations, achieving greater experimental control than when testing on CI subjects. However, it remains an open question whether advancements in algorithms developed on NH subjects using a simulator will necessarily improve performance in CI users. This study assessed the similarity in vowel identification of CI subjects and NH subjects using an 8-channel noise-band vocoder simulator configured to match input and output frequencies or to mimic output after a basalward shift of input frequencies. Under each stimulus condition, NH subjects performed the task both with and without feedback/training. Similarity of NH subjects to CI users was evaluated using correct identification rates and information theoretic approaches. Feedback/training produced higher rates of correct identification, as expected, but also resulted in error patterns that were closer to those of the CI users. Further evaluation remains necessary to determine how patterns of confusion at the token level are affected by the various parameters in CI simulators, providing insight into how a true CI simulation may be developed to facilitate more rapid prototyping and testing of novel CI signal processing and electrical stimulation strategies.

Rhythm measures and dimensions of durational variation in speech

Patterns of durational variation were examined by applying 15 previously published rhythm measures to a large corpus of speech from five languages. In order to achieve consistent segmentation across all languages, an automatic speech recognition system was developed to divide the waveforms into consonantal and vocalic regions. The resulting duration measurements rest strictly on acoustic criteria. Machine classification showed that rhythm measures could separate languages at rates above chance. Within-language variability in rhythm measures, however, was large and comparable to that between languages. Therefore, different languages could not be identified reliably from single paragraphs. In experiments separating pairs of languages, a rhythm measure that was relatively successful at separating one pair often performed very poorly on another pair: there was no broadly successful rhythm measure. Separation of all five languages at once required a combination of three rhythm measures. Many triplets were about equally effective, but the confusion patterns between languages varied with the choice of rhythm measures.

Facial emotion labeling in language impaired children

ABSTRACTThe few studies that have investigated emotion labeling in children with specific language impairment (SLI) have generally focused on global identification performances and appear contradictory. The current study is a fine-grained examination of how children with SLI and typical peers differ in the accuracy of their emotional lexicon use. Children underwent a free labeling task of five basic emotions expressed by still face photographs. Results revealed that children with SLI were less accurate in their label use than typical children. However, pattern of confusions between the two groups differed only by a confusion between sadness and anger displayed by the SLI group. It is argued that this emotion labeling deficiency may rely on semantic fields overlap.

A TS-Type Maximizing-Discriminability-Based Recurrent Fuzzy Network for Classification Problems

This work proposes a Takagi-Sugeno (TS)-type maximizing-discriminability-based recurrent fuzzy network (MDRFN) that can classify highly confusable patterns. The discriminative capability plays a significant role in determining classification performance. To increase the discriminative capability, the proposed MDRFN considers minimum classification error (MCE) and minimum training error (MTE). In MCE, the weights are updated by maximizing the discrimination among different classes. In MTE, the parameter learning adopts the gradient-descent method to reduce the cost function. Therefore, the novelty of MDRFN is that it not only minimizes the cost function but maximizes the discriminative capability as well. The effectiveness of the proposed MDRFN is demonstrated by three temporal classification problems. In the experiments, other RFNs, including the singleton-type recurrent neural fuzzy network (SRNFN), TS-type RFN (TRFN), and simple RFN (SRFN), are compared. Analysis results indicate that the proposed MDRFN exhibits excellent classification performance.

Vowel confusion patterns in adults during initial 4 years of implant use

This study investigated adult cochlear implant users' (n = 39) vowel recognition and confusions by an open-set syllable test during 4 years of implant use, in a prospective repeated-measures design. Subjects' responses were coded for phoneme errors and estimated by the generalized mixed model. Improvement in overall vowel recognition was highest during the first 6 months, showing statistically significant change until 4 years, especially for the mediocre performers. The best performers improved statistically significantly until 18 months. The poorest performers improved until 12 months and exhibited more vowel confusions. No differences were found in overall vowel recognition between Nucleus24M/24R and Med-ElC40+ device users (matched comparison), but certain vowels showed statistically significant differences. Vowel confusions between adjacent vowels were evident, probably due to the implant users' inability to discriminate formant frequencies. Vowel confusions were also dominated by vowels whose average F1 and/or F2 frequencies were higher than the target vowel, indicating a basalward shift in the confusions.

A multimodal approach to emotion recognition ability in autism spectrum disorders

Autism spectrum disorders (ASD) are characterised by social and communication difficulties in day-to-day life, including problems in recognising emotions. However, experimental investigations of emotion recognition ability in ASD have been equivocal, hampered by small sample sizes, narrow IQ range and over-focus on the visual modality. We tested 99 adolescents (mean age 15;6 years, mean IQ 85) with an ASD and 57 adolescents without an ASD (mean age 15;6 years, mean IQ 88) on a facial emotion recognition task and two vocal emotion recognition tasks (one verbal; one non-verbal). Recognition of happiness, sadness, fear, anger, surprise and disgust were tested. Using structural equation modelling, we conceptualised emotion recognition ability as a multimodal construct, measured by the three tasks. We examined how the mean levels of recognition of the six emotions differed by group (ASD vs. non-ASD) and IQ (≥ 80 vs. < 80). We found no evidence of a fundamental emotion recognition deficit in the ASD group and analysis of error patterns suggested that the ASD group were vulnerable to the same pattern of confusions between emotions as the non-ASD group. However, recognition ability was significantly impaired in the ASD group for surprise. IQ had a strong and significant effect on performance for the recognition of all six emotions, with higher IQ adolescents outperforming lower IQ adolescents. The findings do not suggest a fundamental difficulty with the recognition of basic emotions in adolescents with ASD.

Visual accessibility of ramps and steps

The visual accessibility of a space refers to the effectiveness with which vision can be used to travel safely through the space. For people with low vision, the detection of steps and ramps is an important component of visual accessibility. We used ramps and steps as visual targets to examine the interacting effects of lighting, object geometry, contrast, viewing distance, and spatial resolution. Wooden staging was used to construct a sidewalk with transitions to ramps or steps. Forty-eight normally sighted subjects viewed the sidewalk monocularly through acuity-reducing goggles and made recognition judgments about the presence of the ramps or steps. The effects of variation in lighting were milder than expected. Performance declined for the largest viewing distance but exhibited a surprising reversal for nearer viewing. Of relevance to pedestrian safety, the step up was more visible than the step down. We developed a probabilistic cue model to explain the pattern of target confusions. Cues determined by discontinuities in the edge contours of the sidewalk at the transition to the targets were vulnerable to changes in viewing conditions. Cues associated with the height in the picture plane of the targets were more robust.

The effects of frequency-place shift on consonant confusion in cochlear implant simulations.

The effects of frequency-place shift on consonant recognition and confusion matrices were examined. Frequency-place shift was manipulated using a noise-excited vocoder with 4 to 16 channels. In the vocoder processing, the location of the most apical carrier band varied from the matched condition (i.e., 28 mm from the base of the cochlear) to a basal shift (i.e., 22 mm from the base) in a step size of 1 mm. Ten normal-hearing subjects participated in the 20-alternative forced-choice test, where the consonants were presented in a /Ca/ context. Shift of 3 mm or more caused the consonant recognition scores to decrease significantly. The effects of spectral resolution disappeared when the amount of shift reached >or=3 mm. Information transmitted for voicing and place of articulation varied with spectral shift and spectral resolution, while information transmitted for manner was affected only by spectral shift but not spectral resolution. Spectral shift has shown specific effects on the confusion patterns of the consonants. The direction of errors reversed as spectral shift increased and the patterns of reversal were consistent across channel conditions. Overall, transmission of the consonant features can be accounted for by the acoustic features of the speech signal.

Recognition of accented English in quiet by younger normal-hearing listeners and older listeners with normal-hearing and hearing loss

This investigation examined the effects of listener age and hearing loss on recognition of accented speech. Speech materials were isolated English words and sentences that featured phonemes that are often mispronounced by non-native speakers of English whose first language is Spanish. These stimuli were recorded by a native speaker of English and two non-native speakers of English: one with a mild accent and one with a moderate accent. The stimuli were presented in quiet to younger and older adults with normal-hearing and older adults with hearing loss. Analysis of percent correct recognition scores showed that all listeners performed more poorly with increasing accent, and older listeners with hearing loss performed more poorly than the younger and older normal-hearing listeners in all accent conditions. Context and age effects were minimal. Consonant confusion patterns in the moderate accent condition showed that error patterns of all listeners reflected temporal alterations with accented speech, with major errors of word-final consonant voicing in stops and fricatives, and word-initial fricatives.

Characterization of human tactile pattern recognition performance at different ages

This study examined tactile pattern recognition performance in human observers (N = 44) in the context of a letter recognition task at the fingertip. Participants were recruited from three different age groups (youth, n = 17; young adults, n = 14; seniors, n = 13) to examine age-related differences in performance. The influence of gender (males vs females) and hand (right vs left) was also examined. Performance was characterized in terms of both response accuracy and associated response times (RTs). Patterns of confusion between letters were also examined. Results showed that age was the most important factor in determining the capacity of our participants to perform fast and accurate pattern recognition. In this respect, younger participants (i.e., youth and young adults) clearly outperformed seniors by showing not only better accuracy and less confusion but also 2–3 times faster RT. By comparison, the combined influence of “hand” and “gender” on recognition performance was only marginal. These results indicate that the ability to perform complex tactile pattern recognition is already well established in youth 10–14 years of age with only minor refinements occurring later in early adulthood. With advancing age, such ability becomes far less efficient, as judged by the drastic increase in RT observed in seniors, in spite of a relatively good accuracy. This suggests that alterations not only at the peripheral receptor level but also at the central processing level might play an important role in limiting the ability of seniors to perform fast and efficient pattern recognition at the fingertip.

Identification of Acoustically Modified Mandarin Tones by Non-native Listeners

This study investigated identification of fragmented Mandarin tones by non-native listeners. Monosyllabic Mandarin words were digitally processed to generate intact, silent-center, center-only, and onset-only syllables. The syllables were recorded with two carrier phrases such that the offset of the carrier tone and the onset of the target tone were either continuous or discontinuous in fundamental frequency (F0). The syllables were presented with an original carrier phrase, excised from the carrier phrase, or excised and cross-spliced with another carrier phrase. Response accuracy and reaction time were measured, and tone confusion patterns were analyzed. Overall, tone identification varied as a function of modification and tone. Intact and center-only syllables were identified more accurately than silent-center and onset-only syllables. Tone 2 was consistently the most challenging tone to identify. Although the performance level of the third-year students approached that of native listeners reported in Lee, Tao, and Bond (2008), the non-native listeners did not show evidence of using coarticulatory information. Nonetheless, the continuity or discontinuity in F0 between the carrier and target tones did affect tone identification, suggesting the influence of context in non-native tone identification.

Children’s recognition of American English consonants in noise

In contrast to the availability of consonant confusion studies with adults, to date, no investigators have compared children's consonant confusion patterns in noise to those of adults in a single study. To examine whether children's error patterns are similar to those of adults, three groups of children (24 each in 4-5, 6-7, and 8-9 yrs. old) and 24 adult native speakers of American English (AE) performed a recognition task for 15 AE consonants in /ɑ/-consonant-/ɑ/ nonsense syllables presented in a background of speech-shaped noise. Three signal-to-noise ratios (SNR: 0, +5, and +10 dB) were used. Although the performance improved as a function of age, the overall consonant recognition accuracy as a function of SNR improved at a similar rate for all groups. Detailed analyses using phonetic features (manner, place, and voicing) revealed that stop consonants were the most problematic for all groups. In addition, for the younger children, front consonants presented in the 0 dB SNR condition were more error prone than others. These results suggested that children's use of phonetic cues do not develop at the same rate for all phonetic features.

The use of confusion patterns to evaluate the neural basis for concurrent vowel identification.

Perceptual studies of concurrent vowel identification suggest that listeners with normal hearing (NH) are better able to use differences in fundamental frequency (F0) than listeners with sensorineural hearing loss (SNHL). However, the neural basis for this difference remains unknown. The present study sought to validate a neural model of concurrent vowel identification based on specific confusion patterns made by NH listeners. A standard set of five vowels was used. In each concurrent vowel pair, vowel 1 had F0 = 100 Hz, while the F0 of vowel 2 was varied between 100–126 Hz (4 semitones). NH listeners made similar confusions across all F0 differences, with a reduction in occurrence as F0 difference increased. F0 benefit varied significantly across concurrent vowel pairs. A phenomenological auditory-nerve model was cascaded with F0 segregation algorithms to predict the perceptual observations. Neural predictions showed similar confusion patterns to the perceptual data for many (but not all) concurrent vowel pairs. Validating a NH physiological model with specific confusion patterns will allow the model to be used to predict effects of specific physiological changes related to SNHL, which may be useful for improving cochlear-implant and hearing-aid signal processing strategies. [Work supported by Purdue University.]

A mathematical model of vowel identification by users of cochlear implants.

A simple mathematical model is presented that predicts vowel identification by cochlear implant users based on these listeners' resolving power for the mean locations of first, second, and/or third formant energies along the implanted electrode array. This psychophysically based model provides hypotheses about the mechanism cochlear implant users employ to encode and process the input auditory signal to extract information relevant for identifying steady-state vowels. Using one free parameter, the model predicts most of the patterns of vowel confusions made by users of different cochlear implant devices and stimulation strategies, and who show widely different levels of speech perception (from near chance to near perfect). Furthermore, the model can predict results from the literature, such as Skinner, et al. [(1995). Ann. Otol. Rhinol. Laryngol. 104, 307-311] frequency mapping study, and the general trend in the vowel results of Zeng and Galvin's [(1999). Ear Hear. 20, 60-74] studies of output electrical dynamic range reduction. The implementation of the model presented here is specific to vowel identification by cochlear implant users, but the framework of the model is more general. Computational models such as the one presented here can be useful for advancing knowledge about speech perception in hearing impaired populations, and for providing a guide for clinical research and clinical practice.

A Maximizing Discriminability Based Self-Organizing Fuzzy Network for Classi¿cation Problems

A maximizing-discriminability-based self-organizing fuzzy network (MDSOFN) that can classify highly confusable patterns is proposed in this paper. The underlying notion of the proposed MDSOFN is to split the generation of fuzzy rules into linear discriminant analysis (LDA) and Gaussian mixture model (GMM). In LDA, the weights are updated by seeking directions that are efficient for discrimination. In GMM, parameter learning adopts the gradient-descent method to reduce the cost function. Since LDA-derived fuzzy rules increase the discriminative capability among different classes, the proposed MDSOFN can classify highly confusable patterns. The effectiveness of the proposed MDSOFN is demonstrated by two classification problems. A detailed comparative performance analysis for the fuzzy networks using LDA, principal component analysis (PCA), and the support vector machine (SVM), with various noise types, is presented. Experimental results and theoretical analysis indicate that the LDA-derived fuzzy network performs better than the PCA-based fuzzy network and the SVM-based fuzzy network.