Conversational Co-Speech Gesture Generation via Modeling Dialog Intention, Emotion, and Context with Diffusion Models | Synapse