CaR-FOREST: Joint Classification-Regression Decision Forests for Overlapping Audio Event Detection
This work addresses overlapping audio event detection for audio processing applications, presenting an incremental improvement with mixed results.
The paper tackled overlapping audio event detection by proposing CaR-FOREST, a joint classification-regression decision forest method, which significantly outperformed the baseline on Task2 but was inferior on Task3 in the DCASE 2016 challenge.
This report describes our submissions to Task2 and Task3 of the DCASE 2016 challenge. The systems aim at dealing with the detection of overlapping audio events in continuous streams, where the detectors are based on random decision forests. The proposed forests are jointly trained for classification and regression simultaneously. Initially, the training is classification-oriented to encourage the trees to select discriminative features from overlapping mixtures to separate positive audio segments from the negative ones. The regression phase is then carried out to let the positive audio segments vote for the event onsets and offsets, and therefore model the temporal structure of audio events. One random decision forest is specifically trained for each event category of interest. Experimental results on the development data show that our systems significantly outperform the baseline on the Task2 evaluation while they are inferior to the baseline in the Task3 evaluation.