SAM-GCNN: A Gated Convolutional Neural Network with Segment-Level Attention Mechanism for Home Activity Monitoring

• DOI: 10.48550/arxiv.1810.03986
• Publication Date: 2018-01-01
• Authors: Yuhan Shen; Kexin He; Wei-Qiang Zhang
• Keywords: Baseline (sea)

In this paper, we propose a method for home activity monitoring. We demonstrate our model on dataset of Detection and Classification Acoustic Scenes Events (DCASE) 2018 Challenge Task 5. This task aims to classify multi-channel audios into one the provided pre-defined classes. All these classes are daily activities performed in environment. To tackle task, gated convolutional neural network with segment-level attention mechanism (SAM-GCNN). The proposed framework is two auxiliary modules: mechanism. Furthermore, adopted ensemble enhance capability generalization model. evaluated work development DCASE 5 achieved competitive performance, macro-averaged F-1 score increasing from 83.76% 89.33%, compared baseline system.