Leveraging Speaker Embeddings with Adversarial Multi-task Learning for Age Group Classification
This work addresses age group classification from speech, which is an incremental improvement for applications like biometrics or human-computer interaction.
The paper tackled the problem of age group classification from speech by using speaker embeddings from adversarial multi-task learning to reduce domain discrepancies, achieving improved performance verified on the VoxCeleb Enrichment dataset.
Recently, researchers have utilized neural network-based speaker embedding techniques in speaker-recognition tasks to identify speakers accurately. However, speaker-discriminative embeddings do not always represent speech features such as age group well. In an embedding model that has been highly trained to capture speaker traits, the task of age group classification is closer to speech information leakage. Hence, to improve age group classification performance, we consider the use of speaker-discriminative embeddings derived from adversarial multi-task learning to align features and reduce the domain discrepancy in age subgroups. In addition, we investigated different types of speaker embeddings to learn and generalize the domain-invariant representations for age groups. Experimental results on the VoxCeleb Enrichment dataset verify the effectiveness of our proposed adaptive adversarial network in multi-objective scenarios and leveraging speaker embeddings for the domain adaptation task.