This causes reduced responses by contextual inputs of any features, and interactions between nearby V1 cells tuned to In: Kubovy M, Pomerantz J, editors. If you have any supplemental information about the identity listed here, please click in this box to go to the contribution form. Thank you in advance! Person. The relationship between the behavior of single-celled organisms and cognition in higher animals is explored. Recent research and theory in bacterial.
|Published (Last):||21 January 2006|
|PDF File Size:||15.57 Mb|
|ePub File Size:||17.57 Mb|
|Price:||Free* [*Free Regsitration Required]|
Conceived and designed the experiments: The dny to inkubovh sudden changes in the environment is critical for survival. Importantly, these scenes lack semantic attributes, which may have confounded previous studies, thus allowing us to probe low-level processes involved in auditory change perception.
Listeners are remarkably tuned to object appearance: In contrast, listeners have difficulty detecting disappearing objects, even in small scenes: We also measured change detection performance when a noise or silent gap was inserted at the time of change or when the scene was interrupted by a distractor that occurred at the time of change but did not mask any scene elements. Gaps adversely affected the processing of item appearance but not disappearance. However, distractors reduced both appearance and disappearance detection.
Together, our results inkuovy a role for neural adaptation and sensitivity to transients in the process of auditory change detection, similar to what has been demonstrated for visual change detection. Importantly, listeners consistently performed better for item addition relative to deletion across all scene interruptions used, suggesting a robust perceptual representation of item appearance.
The ability to detect and quickly respond to new unkubovy in the environment is critical to an organism’s struggle for survival. The issue of sensitivity to change wny been a topic of intense investigation in vision .
Conversely, when local, change-related, transients are masked, e. In contrast, the factors that affect listeners’ ability to detect the appearance or disappearance of objects within busy acoustic scenes comprised of multiple concurrent sources remain poorly understood. This is despite the fact that sound is often what alerts us to important changes around us: Hearing is sensitive to a much wider space than the other senses and in many cases we hear a change before we inkugovy it E.
We are aware of only a handful of studies that examined listeners’ sensitivity to change in scene contents  — . Pavani and Turatto  further demonstrated no difference between conditions where the pre- and post-change scenes were contiguous or separated by silent or noise-filled gaps. The explanation offered to this set of findings is that rather than low-level sensory mechanisms, auditory change detection fundamentally relies on limited-capacity acoustic memory  and is not sng, in that it requires directed attention suggesting, rather surprisingly, that the auditory system may not be as sensitive to change as previously assumed.
However, this interpretation is confounded by the use of easily identifiable natural sounds. The use of natural sounds in the laboratory poses several problems.
First, it is dny to control their physical parameters indeed they were not controlled in previous workand thus difficult to relate specific effects to underlying stimulus properties. For example, because the previously used sounds ibkubovy in frequency, inter-element masking grew as scenes became more populated and this might have contributed to the observed deterioration of performance with growing scene size.
Second, the sounds were familiar and associated with semantic labels in many cases subjects were explicitly encouraged and trained to name the sounds, e. The performance limits may consequently reflect limits of general working memory rather than a specific auditory change detection system see also . Here we developed a new paradigm for studying auditory change processing. Each source is characterized by a different frequency, and is furthermore modulated at a distinct amplitude modulation Sng rate to ensure that sources are perceived as separate, distinguishable inkbovy.
The modulation mimics temporal properties found across many natural sounds e. Dashed lines show the nominal change time. sjy
ISNI Gamil’ton, Lorel ()
Channels are smoothed to obtain a temporal resolution similar to the Equivalent Rectangular Inkubovj .
Experimental procedures were approved by the research ethics committee snt University College London, and written informed consent was obtained from each participant. Subjects were paid for their participation.
About 10 paid subjects specific information detailed below took part in each experiment each experiment consisted of a different group of subjects but many participated in more than one. All reported normal hearing and no history of neurological disorder.
The vast majority were not musically trained. Scenes comprised multiple components, each consisting of a pure tone modulated by a square wave.
Square wave steps were shaped by 3 ms raised-cosine ramps. Snny component had a unique frequency and AM rate. Component AM rates were randomly drawn from a pool of 15 fixed values between 3 and 35 Hz random phase.
Unless otherwise specified, scenes consisted of 4, 8, or 14 components. Stimulus duration varied randomly between and ms in steps of ms.
Web pages: What can you see in a single fixation?
The timing of change varied randomly between and ms post scene onset with the constraint that changed components were added inlubovy deleted with zero phase: These were then presented in random order or blocked according to change injubovy during the experiment.
Stimulus presentation was controlled using the Cogent software http: Stimulus conditions specific to each experiment are described below. Experimental sessions lasted between 1. Subjects were instructed to fixate at a cross presented on the computer screen, while performing the task relevant to each experiment. Participants were allowed a short rest between runs. The purpose of Experiment 1 was to determine whether our artificial scenes are suitable for assessing change detection.
We aimed to measure: Experiment 1 used only NC stimuli, associated with a probe in two configurations presented in separate blocks: The probe consisted of a single component amplitude-modulated tone.
The duration of the scene was ms, that of the probe ms, and NC and probe were separated by a ms silence. The probe component was either present or absent within the NC scene, with a probability of 0. Subjects had to determine whether the component was present.
Probe-scene pairs were presented with an inter-trial-interval randomized between and ms. Figure 2 plots the results of Experiment 1. As expected, performance is better for smaller scenes, and furthermore better if the probe precedes the scene. Experiment 1 which tested listener’s ability to judge whether a probe a single AM tone is present within a NC scene.
Plotted are sensitivity scores as inkuboy function of scene size. Error bars are 1 standard error SE. The data also demonstrate that listeners are able, to a iinkubovy extent, to determine, post hoc whether a particular component was part of the scene. The goal of Experiment 2 was to test listeners’ ability to detect sudden changes in the scene, manifested as the appearance or disappearance of an element. We used a broad range of scene sizes 4, 6, 8, 10, inkubovvy, and 14 components.
All stimuli were presented in random order not blocked according to change type with an inter-stimulus-interval ISI between to ms. Subjects were required to press a snh once they detected a change in the scene they were not required to determine the type of change. Results are shown in Figure 3. There is a remarkable difference in performance for CA versus CD stimuli: The interaction suggests that performance decreased with scene size inkuvovy CD trials, whereas scene size had no effect on CA stimuli.
Overall, listeners performed substantially better and faster on CA relative to CD stimuli. Error bars are 1 SE. The addition of a component in a CA scene is associated with a slight loudness increase, while the deletion of a component in a CD scene results in a slight loudness decrease. It could well be that subjects were solving the change detection task by using loudness change as a cue.
Indeed, it has been established that listeners are more sensitive to loudness increments than decrements e. In the following experiment we evaluated the extent to which sensitivity to loudness change may have played a role in listeners’ performance.
Specifically, we introduced prominent at least 6 dB loudness changes to all signals CA, CD, and NC; at the nominal time of changewhich were large enough to mask any loudness-related cues involved in item addition or deletion estimated to be at most 6 dB, in scenes with 4 zny, and lower for larger scenes.
Stimuli in Experiment 3 a were identical to those in Experiment 2 except that a loudness change in the form of a single upwards or downwards step in amplitude was introduced in all stimuli, at the nominal change time. We used a variety of step sizes in order to make the magnitude as well as the direction of loudness change increase vs.
This would encourage the listeners to ignore the loudness change and focus on the change in scene content. However the loudness of some components could be perceived as larger because of unequal sensitivity to frequency. Setting the smallest step size at 6 dB is a conservative estimate designed to address this syn. Subjects were instructed to ignore the loudness changes and focus on detecting a change in the inkubpvy of the scene.
This is expected because the addition of the loudness change, which subjects are required to learn to ignore, makes the change detection task harder. Nevertheless the pattern of results is similar to that of Experiment 2. Importantly, the snj performance on CA vs. A loudness change in the form of a single upwards or downwards step in amplitude was introduced in all stimuli, at the nominal change time.
A, B show inkibovy from Experiment 3 A randomized presentation. The interaction suggests that the scene size effect is stronger for CD than CA trials. Overall, the statistical tests therefore confirm that listeners performed significantly better and faster in detecting inkuboovy rather than disappearing components.
This suggests that the asymmetry between CA and CD nsy not based on sensitivity to loudness change cues. In Experiment 4 we assessed participants’ ability to identify the appearing or disappearing components.
That is, we sought to measure whether, in addition to detecting that some change has occurred, listeners have ijkubovy to more detailed information about what changed. To do this, a secondary probe recognition task was administered during trials sy which a change was detected.
Two of the participants were musically trained. The stimuli in Experiment 4 A were identical to those in Experiment 2 except that each time the participant pressed a button to indicate that change was detected, the scene was interrupted and a probe ms amplitude modulated pure tone was presented ms later. The probe was either the component inkuubovy with the change or one of the other components present in the scene, with equal probability.
Subjects were required to judge whether the probe was identical to the changed component. They were encouraged to focus primarily on the change detection task, inkuovy guess if unsure about the probe task. So as to not inadvertently provide feedback, inkubogy were also presented after false positives response to a NC stimulusbut those trials were not analysed.