walter_slides

walter_slides - 128-sample (~8ms) segments taken at random...

Info iconThis preview shows pages 1–19. Sign up to view the full content.

View Full Document Right Arrow Icon
Efficient Coding of Natural Sounds Michael S. Lewicki
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Efficient Coding ±The hypothesis is that sensory relays recode sensory messages so that their redundancy is reduced but comparatively little information is lost± - H. B. Barlow Exploit statistical regularities in the environment to streamline the neural representation of stimuli.
Background image of page 2
Efficient Coding of Images Olshausen ± Field 1996:
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Efficient Coding of Images Bell ± Sejnowski 1997:
Background image of page 4
Model of Auditory Response a i is the response (of neuron i) x is the input stimulus h i is the filter (±receptive field± of neuron i) convolve filter with input to get response
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Auditory Nerve Responses
Background image of page 6
Data Environmental Sounds Twigs snapping, rain, fire, streams Vocalizations Animal sounds from the neotropical rainforest Speech Samples from the TIMIT dataset
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 14
Background image of page 15

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 16
Background image of page 17

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 18
Background image of page 19
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 128-sample (~8ms) segments taken at random from the sets Filters Independent components found using the gradient of the probability of the signals given the components Not intended to be biologically plausible Results - Environmental Sounds Results - Vocalization Results - Speech Results - 2 Sound : 1 Vocalization Time/Frequency Tradeoff PCA on Environmental Sounds Sparse Noise and Single Speaker Time-Frequency Analysis Comparison to Physiological Data + Voc o Sp x Es Equalized power spectra Discussion Results provide an explanation for cochlear responses as efficient codes of natural sounds Limitations: Linear Small timescale Not causal filters Resemblance of filters to data shows that efficient codes evolved Why not learned?...
View Full Document

This note was uploaded on 09/15/2011 for the course COGS 1 taught by Professor Lewis during the Spring '08 term at UCSD.

Page1 / 19

walter_slides - 128-sample (~8ms) segments taken at random...

This preview shows document pages 1 - 19. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online