Audio Cue Research Articles

Semantic versus Spatial Audio Cues: Is There a Downside to Semantic Cueing?

An auditory spatial Stroop paradigm was used to examine the effects of semantic and spatial audio cue conflict on accuracy and response time. Participants responded to either the semantic meaning or the spatial location of a directional word, which was either congruent (i.e. the word “right” being presented from the right) or incongruent (i.e. the word “right” being presented from the left). Contrary to our predictions, people responded more accurately to the semantic meaning of the directional words relative to spatial audio cues. An interaction between task type and congruency indicated people found it more difficult to ignore the semantic content of the word when performing the location version of this task relative to ignoring the location in the semantic task. Implications of these results for use in driver collision-avoidance warning systems are discussed. The current results indicate that performance benefits gained by the use of directional words in combination with spatial audio may be offset by greater detriment to performance when the warning system presents unreliable or incongruent semantic information.

Associative learning by cattle to enable effective and ethical virtual fences

An ability of cattle to readily associate a non-aversive audio cue (conditioned stimulus) with an aversive but non-noxious electric shock (unconditioned stimulus) should enable virtual fences to control cattle in an ethical manner similar to conventional electric fencing. The first study was conducted to identify an effective audio cue. Audio (784 Hz tone) and shock (600 V, 250 mW) stimuli were delivered by remote control to GPS collars on five heifers to prevent access to an exclusion zone surrounding a feed trough. An audio cue was administered when the animal entered the exclusion zone, followed by a shock if the animal continued to proceed. There was an increase in the proportion of heifers responding favourably to the audio cue by turning, backing up or stopping in sessions 3 and 4 (73%) compared with sessions 1 and 2 (44%). This indicated that cattle associated the audio cue with the electric shock and learnt to avoid the trough. The main study examined whether cattle location can be controlled by an audio conditioned stimulus without the presence of a visual cue. Weeks 1 and 2 tested heifers’ learning of the association between an audio conditioned stimulus and an electric shock reinforcer. In week 3, the effect of dispensing with the conditioned stimulus was tested. Heifers were allocated to two treatments ( n = 11). Treatment 1 received an audio cue and an electric shock on exclusion zone entry, as in the first 2 weeks. Treatment 2 received no audio cue and only an electric shock on exclusion zone entry. There was a difference in the behaviours shown in response to both the audio and shock stimuli between weeks 1 and 2, with more heifers turning in response to the audio cue in week 2 than in week 1. When the virtual fence was moved in week 2, 80% of animals ignored the first audio cue, but the proportion failing to respond to the second audio dropped to 46%, indicating that animals had learnt to avoid the electric shock by responding to the audio cue alone to remain within the virtual fence boundary. In week 3, heifers received significantly fewer shocks when a conditioned stimulus was used. There were no differences between treatments in scores for effectiveness of the fence, appropriateness of the stimulus and adverse responses. This study demonstrates that the appropriate use of an audio cue is an effective conditioned stimulus for virtual fencing of cattle.

Size variance of motor evoked potential at initiation of voluntary contraction in palsy of conversion disorder

To investigate the efficacy of transcranial magnetic stimulation (TMS) with a cue signal for the objective diagnosis of palsy of conversion disorder (CD). Ten patients with palsy of CD, nine with amyotrophic lateral sclerosis (ALS), and eight control subjects were examined. Motor evoked potential (MEP) was recorded from the abductor digiti minimi muscle under three conditions: at rest, during tonic contraction, and with an audio cue signal. In the cue signal paradigm, subjects were asked to perform ramp-and-hold contraction in response to a cue signal. MEP size increased in the cue signal paradigm in both controls and patients with ALS, but was not obvious in some CD patients. This was likely due to variance among trials in the cue signal paradigm in each CD patient. The coefficients of variance (CV) among trials in the cue signal paradigm were 15 +/- 4.3 in controls, 25 +/- 11 in ALS, and 70 +/- 40 in CD. CV of MEP size with cue signal was larger in CD than in controls (P < 0.005) and in ALS patients (P < 0.01). The size variance among MEP trials with the cue signal is a supportive parameter for the diagnosis of CD.

Stepping Backward Can Improve Sprint Performance Over Short Distances

The use of a backward (false) step to initiate forward movement has been regarded as an inferior starting technique and detrimental to sprinting performance over short distances as it requires additional time to be completed, but little evidence exists to support or refute this claim. Therefore, we recruited 27 men to examine the temporal differences among three standing starts that employed either a step forward (F) or a step backward (B) to initiate movement. An audio cue was used to mark the commencement of each start and to activate the subsequent timing gates. Three trials of each starting style were performed, and movement (0 m), 2.5 m, and 5 m times were recorded. Despite similar performances to the first timing gate (0.80 and 0.81 s for F and B, respectively), utilizing a step forward to initiate movement resulted in significantly slower sprint times to both 2.5 and 5 m (6.4% and 5.3%, respectively). Furthermore, when the movement times were removed and performances were compared between gates 1 and 2, and 2 and 3, all significant differences were seen before reaching a distance of only 2.5 m. The results from this investigation question the advocacy of removing the false step to improve an athlete's sprint performance over short distances. In fact, if the distance to be traveled is as little as 0.5 m in the forward direction, adopting a starting technique in which a step backward is employed may result in superior performance.

Video technology monitors hand washing.

A new device which helps people with dementia to wash their hands has been developed using video technology. It works by filming people with dementia washing their hands. If they forget which stage they are at it will prompt them with an audio or visual cue.

Reproducibility of Functional MR Imaging: Preliminary Results of Prospective Multi-institutional Study Performed by Biomedical Informatics Research Network

To prospectively investigate the factors--including subject, brain hemisphere, study site, field strength, imaging unit vendor, imaging run, and examination visit--affecting the reproducibility of functional magnetic resonance (MR) imaging activations based on a repeated sensory-motor (SM) task. The institutional review boards of all participating sites approved this HIPAA-compliant study. All subjects gave informed consent. Functional MR imaging data were repeatedly acquired from five healthy men aged 20-29 years who performed the same SM task at 10 sites. Five 1.5-T MR imaging units, four 3.0-T units, and one 4.0-T unit were used. The subjects performed bilateral finger tapping on button boxes with a 3-Hz audio cue and a reversing checkerboard. In a block design, 15-second epochs of alternating baseline and tasks yielded 85 acquisitions per run. Functional MR images were acquired with block-design echo-planar or spiral gradient-echo sequences. Brain activation maps standardized in a unit-sphere for the left and right hemispheres of each subject were constructed. Areas under the receiver operating characteristic curve, intraclass correlation coefficients, multiple regression analysis, and paired Student t tests were used for statistical analyses. Significant factors were subject (P < .005), k-space (P < .005), and field strength (P = .02) for sensitivity and subject (P = .03) and k-space (P = .05) for specificity. At 1.5-T MR imaging, mean sensitivities ranged from 7% to 32% and mean specificities were higher than 99%. At 3.0 T, mean sensitivities and specificities ranged from 42% to 85% and from 96% to 99%, respectively. At 4.0 T, mean sensitivities and specificities ranged from 41% to 73% and from 95% to 99%, respectively. Mean areas under the receiver operating characteristic curve (+/- their standard errors) were 0.77 +/- 0.05 at 1.5 T, 0.90 +/- 0.09 at 3.0 T, and 0.95 +/- 0.02 at 4.0 T, with significant differences between the 1.5- and 3.0-T examinations and between the 1.5- and 4.0-T examinations (P < .01 for both comparisons). Intraclass correlation coefficients ranged from 0.49 to 0.71. MR imaging at 3.0- and 4.0-T yielded higher reproducibility across sites and significantly better results than 1.5-T imaging. The effects of subject, k-space, and field strength on examination reproducibility were significant.

BEHAVIORAL TRAINING AND HYDRAULIC CHUTE RESTRAINT ENABLES HANDLING OF ELAND ANTELOPE (TAUROTRAGUS ORYX) WITHOUT GENERAL ANESTHESIA

Difficulties and risks associated with restraining large nondomestic ungulates are limiting factors toward developing and applying assisted reproductive technologies, such as artificial insemination and embryo transfer. In this study on 10 female eland (Taurotragus oryx), we evaluated the use of behavioral training and handling handling in a hydraulic chute to perform transvaginal ultrasound-guided oocyte retrieval and other clinical procedures. Nine females were conditioned to associate specific sound cues with food treats. The interval from the audio cue until acceptance of handheld treats varied among females (1.8-58.3 min). Animals also differed in their response to training for voluntary entry into the chute. Handling eland for oocyte retrieval in the hydraulic chute required xylazine sedation. During sedation and handling, eland undergoing oocyte retrieval procedures had higher blood glucose levels (14.4 +/- 3.1) than females handled similarly but without oocyte retrieval (9.3 +/- 2.7 mmol/L). Plasma osmotic pressure, hematocrit, and creatine phosphokinase activity were similar between these two groups. Females that were more difficult to train had higher blood glucose levels than the more cooperative animals. Cooperative females had fewer vertical stripes on their sides. More than 40 procedures were conducted without complications or mortality. The combination of behavioral conditioning-training and restraint of sedated eland in a hydraulic chute was a reliable and repeatable method for performing minimally invasive assisted reproductive techniques.

Effects of Supplementing Head-Down Displays With 3-D Audio During Visual Target Acquisition

We investigated the effectiveness of supplementing head-down displays (HDDs) with high-fidelity 3-dimensional (3-D) audio using a flight simulation task in which participants were required to visually acquire the image of a target aircraft. There were 3 conditions: a visual HDD providing azimuth information combined with a nonspatial audio cue, a visual HDD providing azimuth and elevation information combined with a nonspatial audio cue, and a visual HDD providing azimuth information combined with a 3-D audio cue. When 3-D audio was presented, the visual acquisition time was faster, perceived workload was reduced, and perceived situational awareness was improved. This performance improvement was attributed to the fact that participants were often able to perform the task without the need to refer to the HDD.

SUPPORT VECTOR LEARNING FOR GENDER CLASSIFICATION USING AUDIO AND VISUAL CUES

Computer vision systems for monitoring people and collecting valuable demographic information in a social environment is an important research problem. It is expected that such a system will play an increasingly important role in enhancing user's experience and can significantly improve the intelligibility of a human computer interaction (HCI) system. For example, a robust gender classification system can provide a basis for passive surveillance and access to a smart building using demographic information or can provide valuable consumer statistics in a public place. The option of an audio cue in addition to the visual cue promises a robust solution with high accuracy and ease-of-use in human computer interaction systems. This paper investigates gender classification using Support Vector Machines (SVMs). The visual (thumbnail frontal face) and the audio (features from speech data) cues were considered for designing the classifier. Three different representations of the data, namely, raw data, principle component analysis (PCA) and non-negative matrix factorization (NMF) were used for the experimentation with visual signal. For speech, mel-cepstral coefficient and pitch were used for the experimentation. It was found that the best overall classification rates obtained using the SVM for the visual and speech data were 95.31% and 100%, respectively, on data set collected in laboratory environment. The performance of the SVM was compared with two simple classifiers namely, the nearest prototype neighbor and the k-nearest neighbor on all feature sets. It was found that the SVM outperformed the other two classifiers on all datasets. To further understand the robustness issues, the proposed approach has been applied on a large balanced (roughly equal distribution of gender, ethnicity and age group) data-base consisting of 8000 faces collected in real world environment. While, the results are very promising it indicates more to be done to make a statistically meaningful conclusion.

A robust video scene extraction approach to movie content abstraction

AbstractThis research addresses the problem of automatically extracting semantic video scenes from feature films based on multimodal information. A three‐stage scene detection scheme is proposed. First, we use pure visual information to extract a coarse‐level scene structure based on generated shot sinks. Second, audio cue is integrated to refine the scene detection results by considering various kinds of audiovisual scenarios. Finally, we introduce users into this process by allowing them to interactively tune the final results to their own satisfaction. The generated scene structure forms a compact yet meaningful abstraction of the video data, which can help facilitate the content access. Preliminary experiments on integrating multiple media cues for movie scene extraction have yielded encouraging results. © 2004 Wiley Periodicals, Inc. Int J Imaging Syst Technol 13, 236–244, 2003; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ima.10063

An examination of audio and visual cue framing within Australian in‐cinema advertisements

Uses Resnik and Stern’s content analysis criteria to examine audio and visual information of in‐cinema slide advertisements within one regional market in Australia to determine whether two types of cues are compatible or reinforce one another. Suggests that there was extensive information framing for a narrow set of information cues. States that there were also significant differences in the types of audio and visual cues, which might result in conflicting information being communicated or information overload.

Microprocessor-based gait analysis system to retrain Trendelenburg gait.

A microprocessor-based gait analysis system is described that uses two electromyogram (EMG) amplifiers, two foot switches and an audio feedback device to allow the retraining of one type of improper gait, where the hip abductors (gluteus medius muscles) are weak on one side of the body, causing the opposite hip to drop during the swing phase of gait (Trendelenburg gait). As the abnormality is strictly on one side of the body in most people, the circuitry is minimised, as gait can be analysed by only comparing muscle activity in the affected gluteus medius muscle with that in the unaffected gluteus medius muscle, through the EMG. Two foot contact switches are used to help assess timing of the step cycle. If gait is different on the two sides of the body, an audio cue directs the patient to correct the abnormality by increasing activity on the affected side. The device is tested on five patients. Trendelenburg gait is reduced by an average of 29 degrees through the use of the device. The average stride length at the beginning of the study is 0.32 +/- 0.3 m. By the end of the study, the stride length is increased to 0.45 +/- 0.2 m for the entire group of five subjects. The speed of gait has increased from 1.6 +/- 0.4 kmh(-1) to 3.1 +/- 0.5km h(-1).

Ocular Pursuit Responses to Repeated, Single-Cycle Sinusoids Reveal Behavior Compatible With Predictive Pursuit

The link between anticipatory smooth eye movements and prediction in sinusoidal pursuit was investigated by presentation of series of identical, single-cycle, sinusoidal target motion stimuli. Stimuli occurred at randomized intervals (1.2-2.8 s) but were preceded by an audio warning cue 480 ms before each presentation. Cycle period (T) varied from 0.64 to 2.56 s and target displacement from 4 to 20 degrees in separate series. For T </= 1.28 s, responses to the first stimulus of each series exhibited a time delay across the whole cycle (mean = 121 ms for T = 0.8 s). But, in the second and subsequent (steady-state) presentations, anticipatory movements, proportional to target velocity, were made and time delay was significantly reduced (mean = 43 ms for T = 0.8 s). Steady-state time delays were comparable to those evoked during continuous sinusoidal pursuit and less than pursuit reaction time. Even when subjects did not follow the target in the first presentation, they responded to the second presentation with reduced time delay. Throughout the experiments, three types of catch trial (A-C) were introduced. In A, the target failed to appear as expected after the warning cue. Anticipatory smooth movements were initiated, reaching a peak velocity proportional to prior target velocity around 200 ms after expected target onset. In B, the target stopped midway through the cycle. Even if the target remained on and was stationary, the eye movement continued to be driven away from the stationary target with a velocity similar to that of prior responses, reaching a peak velocity that was again proportional to expected target velocity after >/=205 ms. In C, the amplitude of the single sinusoid was unexpectedly increased or decreased. When it decreased, eye velocity throughout the first half-cycle of the response was close to that executed in response to prior stimuli of higher velocity and did not return to an appropriate level for 382-549 ms. Conversely, when amplitude increased, eye velocity remained inappropriately low for the first half-cycle. Results of A and C indicate that subjects are able to use velocity information stored from prior presentations to initiate an oculomotor drive that predominates over visual feedback for the first half-cycle. Results of B indicate that the second part of the cycle is also preprogrammed because it continued despite efforts to suppress it by fixation. The results suggest that initial retinal velocity error information can be sampled, stored, and subsequently replayed as a bi-directional anticipatory pattern of movement that reduces temporal delay and could account for predictive control during sinusoidal pursuit.

The remembered pursuit task: evidence for segregation of timing and velocity storage in predictive oculomotor control

Regular, repeated presentation of identical constant-velocity target motion stimuli (ramps) appears to allow build up of an internal store, release of which can be used to generate anticipatory smooth pursuit prior to subsequent target onset. Here, we examine whether release of the anticipatory response can be controlled by timing cues unrelated to the motion stimulus itself. In experiment 1, the target moved in alternate directions and was exposed for 480 ms as it passed through centre; otherwise subjects were in darkness. Inter-stimulus interval (ISI) was either regular (3.6 s) or randomized (2.7-4.3 s). Presentations were given with or without audio cues that occurred at a constant cue time (CT) prior to target appearance. Even when ISI was randomized, cues could be used to generate anticipatory smooth pursuit. Eye velocity (V100) measured 100 ms after target onset (just prior to visual feedback influence) was greater with cues than without and decreased significantly as CT increased from 240-960 ms. In experiment 2, we assessed the effects of fixation between presentations and eccentricity of target starting position, using unidirectional ramps. The target was visible for 400 ms and started on, ended on or straddled the midline. Subjects held fixation on the midline until an audio cue signalled that preparation for ensuing target appearance could begin. There was no difference in V100 between starting positions or between presence/absence of fixation. In experiment 3, we compared the effects of using audio, visual or tactile cues. All types of cue evoked anticipatory smooth pursuit, but the response to the visual cue was significantly delayed compared with the others. However, V100 was not significantly different between cues. In all experiments, V100 was scaled in proportion to target velocity over the range 12.5-50 degrees /s, showing that this was a truly predictive response. The results provide evidence that timing and velocity storage can be independently controlled through different sensory channels and suggest that the two functions are probably carried out by separate neural mechanisms.

Auditory guided visual search with visual distracters

In order to study the interaction of the auditory and visual systems, a series of experiments has been designed. The first experiment was completed in 1994 and recently published [Perrott et al., Human Factors 38(4), 702–715 (1996)]. This initial experiment explored auditory-aided visual search with a single high-contrast target and a dark background. The results demonstrated a distinct advantage in visual target acquisition with the addition of a correlated directional audio cue. The objective of this experiment was to develop a function describing the effects of directional auditory cueing on searh times in visual fields of varying complexity. Visual conditions included 1, 5, 10, 25, or 50 distracters randomly distributed. The locations of the targets and distracters ranged between ±180° in azimuth and between −75° and +90° in elevation. The three auditory conditions included no auditory cue, a free-field cue, and a virtual cue. In each case, the stimulus was pink noise (20–20 000 Hz). Results will be presented which show the effect of the number of distracters on search times with and without auditory cues. [Work supported by AFOSR.]

Measurement of localization ability using virtual audio cues and visual target acquisition

Visual target detection was used to demonstrate the ability of subjects to localize with a virtual audio display. The testing facility consisted of audio presented over headphones with and without localization cues and a projection screen bearing the visual scene. In each trial the visual target appeared as one pixel at a random location of the screen. The target grew in size until found by the subject. The sound cue came on at threshold level and became progressively louder until the target was found. The audio cue was pink noise while the visual targets consisted of aircraft icons. Subject reaction time was measured from target onset to target detection. As indicated by the preliminary data, the localized audio condition faithfully provided the location of the target resulting in reaction times significantly better than with the nonlocalized audio cue.

A Prototype Visual and Audio Display

A display is described that provides a three-dimensional perspective view with spatially correlated audio. The system is developed around an optics device that projects a three-dimensional perspective view from a CRT to a concave mirror that focuses the energy at an image plane above the mirror. The result is that the objects displayed on the CRT appear to be floating in space. The directional audio is provided from an audio localization cue synthesizer that encodes pinna filtering and an interaural time delay onto an input audio signal. A magnetic head tracker is used to keep the audio images stable. The optical system is presented along with the graphics methods that were used to generate the visual cues. Then the audio localization cue synthesizer is described with emphasis on using this device to provide spatialized audio for use in synthetic environments.

Target detection in reverberation by an echolocating Atlantic bottlenose dolphin (Tursiops truncatus).

An experiment was conducted to determine the ability of an echolocating Atlantic bottlenose dolphin (Tursiops truncatus) to detect targets in the presence of reverberation. Reverberation was induced by a clutter screen consisting of forty-eight 5.1-cm-diam cork spheres spaced 15.2 cm apart and arranged in a rectangular array, located behind the targets. Hollow aluminum cylinders having the same outer diameter and wall thickness and three different lengths were used as targets. The dolphin was trained to station in a hoop, 6 m from the targets, and to echolocate the target upon the reception of an audio cue. Only one of the targets would be presented in a target-present trial. Data were collected on the animal's detection performance as a function of the separation distance between the clutter screen and the targets. The animal's performance for the smallest cylinder varied from 91% to 55% correct as the separation distance decreased from 10.2 to 0 cm. The target strength of the clutter screen and the cylinders were measured both in terms of the energies and the maximum peak-to-peak amplitudes of the incident and reflected echoes. These measurements indicated that when the target-clutter screen separation distance was 0 cm, the dolphin's 50% correct detection occurred at an energy echo-to-reverberation ratio of approximately 0.25 dB and at peak-to-peak echo-to-reverberation ratio of 2.6 dB. The results also indicated that the dolphin's performance varied almost linearly with the echo-to-reverberation ratio. The animal's pulse emissions were monitored by a microprocessor system and the results are presented in terms of the average number of clicks and the average response latency, as a function of the separation distance.

Temporal course of the icon: a summary.

Two experiments were conducted to measure the temporal decay of the peripheral visual memory trace, the icon. In both experiments four paid subjects first learned to map a three-tone by three-location audio cue to the corresponding single location of a nineletter square display briefly presented in a post-stimulus cueing procedure. Thereby temporal specivity of a single target letter to a particulus inter-stimulus-interval was achieved without incidental masking effects from a visual cue. Each subject then generated confusion matrix data for a 10-capital-letter stimulus set at each of six inter-stimulus intervals. The rwo experiments differed in that one used 10 letters as the stimulus set, the other used 10 letters. Both employed inter-srimulus intervals of .05, .IS, .45, .65, .85, and 1.25 sec. A repeated-measures analysis of variance of recognition performance versus interstimulus interval indicated that performance declined with increasing interval for all lines (Flu, = 33.6, df = 5/15, p < .001). However, the performance for bottom-line locations, while initially equal, declined more rapidly than the performance at middle- and top-line-cued locations (PIS[ . LI.D = 2.52, df = 10/102, p < .01). These results suggest that subjects do not, perhaps can not, refrain from recognition processing during the interval, instead processing individually favored display locations until the cue is presented, and that data from post-stimulus cueing studies should be interpreted as including trials in which the cued location was attended prior to cue onset. Performance for bottom-line-cued locations was used to estimate iconic decay unclouded by information processing during the inter-stimulus interval. These data suggested an exponential decay function. Both linear and quadratic components were significant (FI~,.., = 74.62, df = 1/39, p < .001; Fqu.armttc = 6.68, df = 1/39, p < .05). Further analysis focused upon an information-processing model which assumes (1) that the icon is the neural encoding of specific features (Geyer & DeWald, 1973) and (2) that it is the availability of these features which decays exponentially. This model predicted the 12 confusion matrices (six inter-stimulus intervals for each of rwo stimulus sets), with ratios of unexplained residual variance to total variance ranging from .007 to ,024. The feature-list representations of the round letters and those of the rectangular letters included seven features common to both sets (Geyer & DeWald, 1973). Thereby these experiments represented independent estimates of the decay rates of these seven features. The obtained decay rates were related; 7 = .78 (p < .05, single tail). These results suggest modest incrementaI support for the feature lists to capital letters mapping reported by Geyer and DeWald.