Title: Audio Classification and Event Detection Based on Small-size Weakly Labeled Data
Dr. David Anderson, ECE, Chair , Advisor
Dr. Mark Davenport, ECE, Co-Advisor
Dr. Eva Dyer, ECE
Dr. Ghassan AlRegib, ECE
Dr. Elliot Moore, ECE
Dr. Abbas Rashidi, CEE
The objective of this research is to perform audio event detection and classification using small-size weakly labeled data. Although audio event detection has been studied for years, the research on this topic using weakly labeled data is limited. Many sources of multimedia data lack detailed annotation and rather have only high-level meta-data describing the main content of various long segments of the data. In this research, we illustrate a novel framework to perform audio classification when working with such weakly labeled data, especially when dealing with small-size datasets. Traditional approaches to this problem are to use techniques for strongly labeled data and then to deal with the weak nature of the labels via post-processing. In contrast, our approach directly addresses the weakly labeled aspect of the data by classifying longer windows of data based on the clustering behavior of the acoustic features over time. We evaluate our framework using both synthetic datasets and real data and demonstrate that our method works well under both situations. Also, it outperforms other existing methods when using small size datasets.