Figure 1 | Scientific Reports

Figure 1

From: Neuroadaptive modelling for generating images matching perceptual categories

Figure 1

Overview of neuroadaptive generative modelling. Generate: the model G generates digital information based on latent variables \(z_i\). Perceive: a human operator reacts naturally to the generated information represented by the computing system as \(z_i\). Adapt: relevance of the information is inferred from the brain signals of the operator; the relevance guides the generative model, generating new digital information \(G({\hat{z}}_n)\), which matches the operator’s perceptual categories. n is the number of acquired evoked brain signals, here \(n=3\).

Back to article page