CaVIT: An integrated method for image style transfer using parallel CNN and vision transformer.

Integrating gating and learned queries in audiovisual emotion recognition.