Related Experiment Videos
Dichotic speech recognition in noise using reduced spectral cues
Philipos C Loizou1, Arunvijay Mani, Michael F Dorman
1Department of Electrical Engineering, University of Texas at Dallas, P.O. Box 830688, EC 33, Richardson, Texas 75083-0688, USA. loizou@utdallas.edu
The Journal of the Acoustical Society of America
|July 26, 2003
Summary
Spectral resolution impacts how the brain fuses dichotically presented speech. Varying spectral resolution and presentation methods affected speech fusion, particularly for sentences.
Area of Science:
- Auditory Neuroscience
- Speech Perception
- Signal Processing
Background:
- Dichotic speech presentation is influenced by onset timing.
- The role of spectral resolution in dichotic speech fusion is not fully understood.
Purpose of the Study:
- To investigate if spectral resolution influences spectral fusion.
- To examine how splitting spectral information across ears affects speech perception.
Main Methods:
- Speech was synthesized with 6-12 channels at +5 dB S/N.
- Spectral information was split dichotically using odd-even and low-high frequency channels.
- Participants identified speech presented monaurally and dichotically.
Main Results:
- Spectral resolution significantly affected spectral fusion, with sentences being most impacted.
- Lower spectral resolution (6-8 channels) in the low-high condition impaired sentence fusion compared to monaural presentation.
- The odd-even channel split resulted in more accurate sentence identification than the low-high split.
Conclusions:
- Spectral resolution is a key factor in dichotic speech fusion.
- The method of splitting spectral information influences perception.
- Findings have implications for understanding auditory processing and developing speech processing technologies.