white paper -- new science: accent more on providing extra annotations for naturalistic stimuli

[yarikoptic]

per our chat at SfN, to me having rich annotations for naturalistic stimuli datasets (movies) could potentially be the main factor to make them used to MUCH higher degrees of their potential as typically they come with minimal if any annotation, while having very rich collection of dimensions to cover (unless resting state or some controlled experiments).
StudyForrest showed the value of providing additional annotations and projects like https://github.com/neuroscout/neuroscout could potentially employ/make it possible to benefit from richer annotations. Although some methodologies (hyperalignment, "functional connectivity") do not really care about annotations, most of the typical analytics (GLM etc) heavily rely on them. It could be argued that the high dimensionality of the description of natural stimuli paired with its actual absence is what hinders their wider adoption in research practice. Having reach annotations should open roads for development of new analytics on rich neural data - a paradigm shift which is yet to happen (folks still largely do resting state!!!).

Hence, if we do think about New Science, I think it might be valuable to point to that before even jumping into anything fancier cross-modal.

[neuromechanist]

+100. Definitely worth revision with some survey of what is available and what should be developed.