The visome: Using cognitive networks to examine lip-reading errors in English words.

Area of Science:

• Linguistics
• Cognitive Science
• Network Science

Background:

• Lipreading relies on visual cues of spoken words.
• Understanding the visual structure of language is crucial for lipreading.
• Previous research has not extensively utilized network science for visual word analysis.

Purpose of the Study:

• To investigate the influence of visual word network structure on lipreading performance.
• To compare the structural properties of the visome (visual word network) with the phonological network.
• To identify network-based predictors of lipreading errors.

Main Methods:

• Applied network science to analyze the visome and phonological networks.
• Compared network measures at macro, meso, and micro levels.
• Examined psycholinguistic and network structure measures in lipreading error databases.

Main Results:

• Lipreading errors were more frequent than target words.
• Words with later uniqueness points and more 'viseme twins' were lipread less accurately.
• Words with more 'viseme neighbors' also showed reduced lipreading accuracy.
• Errors often occurred within the same network community as the target word.

Conclusions:

• The structure of the visual word network (visome) significantly impacts lipreading accuracy.
• Network analysis provides valuable insights into lipreading challenges.
• Network science offers a promising framework for future lipreading research.