Low-dimensional Embodied Semantics for Music and Language
This addresses the challenge of noisy biological semantics in media artifacts for researchers in cognitive science and AI, though it is incremental in improving existing methods.
The paper tackled the problem of representing shared semantics in music and language by using low-dimensional vector embeddings from joint modeling of multiple human brains, showing that these unsupervised representations outperform high-dimensional fMRI voxel spaces in classification tasks and increase semantic richness.
Embodied cognition states that semantics is encoded in the brain as firing patterns of neural circuits, which are learned according to the statistical structure of human multimodal experience. However, each human brain is idiosyncratically biased, according to its subjective experience history, making this biological semantic machinery noisy with respect to the overall semantics inherent to media artifacts, such as music and language excerpts. We propose to represent shared semantics using low-dimensional vector embeddings by jointly modeling several brains from human subjects. We show these unsupervised efficient representations outperform the original high-dimensional fMRI voxel spaces in proxy music genre and language topic classification tasks. We further show that joint modeling of several subjects increases the semantic richness of the learned latent vector spaces.