DOC PREVIEW
MIT HST 723 - Lecture Notes

This preview shows page 1-2-3-4-25-26-27-51-52-53-54 out of 54 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 54 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Slide 1Slide 2Slide 3Slide 4Slide 5Slide 6Slide 7Slide 8Slide 9Slide 10Slide 11Slide 12Slide 13Slide 14Slide 15Slide 16Slide 17Slide 18Slide 19Slide 20Slide 21Slide 22Slide 23Slide 24Slide 25Slide 26Slide 27Slide 28Slide 29Slide 30Slide 31Slide 32Slide 33Slide 34Slide 35Slide 36Slide 37Slide 38Slide 39Slide 40Slide 41Slide 42Slide 43Slide 44Slide 45Slide 46Slide 47Slide 48Slide 49Slide 50Slide 51Slide 52Slide 53Slide 54>>ITD.mrunning…IC 800Hz40 sp/sec34O azimNeuronApril 16, 2009Bo ZhuHST.723 Spring 2009Theme 3 Paper PresentationApril 1, 2009Background•Reverberation poses sound localization challenge–Interaural Time Difference (ITD) fluctuates from “multiple” sourcesBackground•Reverberation poses sound localization challenge–Interaural Time Difference (ITD) fluctuates from “multiple” sources•Reverberation builds up over time–ITD cues are initially accurate, then degrade over timeBackground•Reverberation poses sound localization challenge–Interaural Time Difference (ITD) fluctuates from “multiple” sources•Reverberation builds up over time–ITD cues are initially accurate, then degrade over time•Cross-correlation model traditionally used for coincidence detection of ITDs–ITD processing initiated in MSO, and elaborated in Inferior Colliculus (IC)Background•Reverberation poses sound localization challenge–Interaural Time Difference (ITD) fluctuates from “multiple” sources•Reverberation builds up over time–ITD cues are initially accurate, then degrade over time•Cross-correlation model traditionally used for coincidence detection of ITDs–ITD processing initiated in MSO, and elaborated in Inferior Colliculus (IC)•Human sound localization is robust in reverberation.Purpose – Examine the ICIf the IC is an important site of ITD sound localization, we need to see how reverberation affects it.Purpose – Examine the ICIf the IC is an important site of ITD sound localization, we need to see how reverberation affects it.What does reverberation do to the IC neurons?Purpose – Examine the ICIf the IC is an important site of ITD sound localization, we need to see how reverberation affects it.What does reverberation do to the IC neurons?Are there mechanisms within the IC itself to handle reverberation? Or does it require higher-level processes?Purpose – Examine the ICIf the IC is an important site of ITD sound localization, we need to see how reverberation affects it.What does reverberation do to the IC neurons?Are there mechanisms within the IC itself to handle reverberation? Or does it require higher-level processes?Does our current model of how the IC works account for reverb robustness?MethodsRecordings of 36 ITD-sensitive IC neurons in anesthetized cats.Binaural stimuli were virtually generated to simulate anechoic, moderate reverb, and strong reverb conditions.Results I Reverb causes Rate CompressionDirectional Sensitivity defined by firing rate range, across whole range of azimuths - (another more rigorous definition in supplement)Results I Reverb causes Rate CompressionDirectional Sensitivity defined by firing rate range, across whole range of azimuths - (another more rigorous definition in supplement)Compression -> Rate Saturation: Inability to correlate firing rates with unique azimuth locations.Sound localization challenge during reverb is already present at the IC neuron level.Results I Reverb causes Rate CompressionDirectional Sensitivity defined by firing rate range, across whole range of azimuths - (another more rigorous definition in supplement)Compression -> Rate Saturation: Inability to correlate firing rates with unique azimuth locations.Sound localization challenge during reverb is already present at the IC neuron level.More reverb -> more rate compressionResults II Cross-correlation modelCross-correlation model was createdfor each single neuron recordedResults II Cross-correlation modelCross-correlation model was createdfor each single neuron recordedModel parameters were adjusted to match anechoic rate-ITD data collected from the actual neuron.Results II Cross-correlation modelCross-correlation model was createdfor each single neuron recordedModel parameters were adjusted to match anechoic rate-ITD data collected from the actual neuron.Reverberant stimuli were input, and model outputted rate-ITD curves.Results II Cross-correlation model poorly predicts directional sensitivity during reverbModel generally behaves well:Similar rate compression evidenced during reverbResults II Cross-correlation model poorly predicts directional sensitivity during reverbModel generally behaves well:Similar rate compression evidenced during reverbModel not accurate during reverb:Accurate prediction would produce points along the y=x diagonal line.Model underestimates performance:Real observed rate ranges (indicating directional sensitivity) are generally higher than the models’.Results II Cross-correlation model poorly predicts directional sensitivity during reverbModel generally behaves well:Similar rate compression evidenced during reverbModel not accurate during reverb:Accurate prediction would produce points along the y=x diagonal line.Model underestimates performance:Real observed rate ranges (indicating directional sensitivity) are generally higher than the models’.This suggests existence of additional reverb robustness in IC neuronsTime Out (pun intended)• So far, we have ignored time effects on rate responses.• All spiking rates have been averaged across the duration of the 400 ms response.Time Out (pun intended)• So far, we have ignored time effects on rate responses.• All spiking rates have been averaged across the duration of the 400 ms response.• ITD degrades as response time progressesTime Out (pun intended)• So far, we have ignored time effects on rate responses.• All spiking rates have been averaged across the duration of the 400 ms response.• ITD degrades as response time progresses • Instead of averaging firing rate over whole response,let’s split response into “early” and “ongoing”Results III Reverb robustness in early IC responseDirectional sensitivity of IC neurons is better early in reverb responseFor this particular unit, we observemore rate compression later in the responseResults III Reverb robustness in early IC responseDirectional sensitivity of IC neurons is better early in reverb responseFor this particular unit, we observemore rate compression later in the


View Full Document

MIT HST 723 - Lecture Notes

Download Lecture Notes
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Lecture Notes and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Lecture Notes 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?