What does this research mean for the field?

Question

Accepted Answer

Adapting the initial layer filters of CNN-based speech activity detection models with a small amount of supervised data significantly improves performance by 10-25% in mismatched acoustic conditions compared to conventional DNN-based systems. Novelty: ClaimNovelty.INCREMENTAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

Analyzing convolutional neural networks for speech activity detection in mismatched acoustic conditions

Discussion

Member takes

Overview

Key Points

Cite This Study