List All Projects List Graduate Projects List Senior Design Projects List REU Projects Other Undergraduate Projects
Project # 51
Sequential versus Hierarchical Grammar Models for Great Tit Vocal Sequence Analysis
| Team Lead: | Sadbhav Kattel |
| Other Team Members: | Abhipsa Guha |
| Faculty Advisor(s): | Dr. Kenny Zhu |
| Department: | CSE |
| Project Type: | REU |
| Title: | Sequential versus Hierarchical Grammar Models for Great Tit Vocal Sequence Analysis |
| Abstract: | Understanding the organizational principles of animal communication provides insight into the evolution of structured signaling and potential precursors to human language. We examine whether great tit (Parus major) vocal sequences are better explained by hierarchical grammar models or sequential Markov-based approaches using 19,395 field recordings with 10 distinct call types. Results show that 3rd-order Markov models (2.36 perplexity) significantly outper form hierarchical approaches (15.71 perplexity), indicating strong local dependencies in call transitions. These findings suggest great tit vocal communication relies on sequential rather than hierarchical principles, contrasting with human language structure. |
| Poster: | View Poster #51 (PDF) |
| Judges Scoring: | Submit Evaluation (judges only) |