While previous gesture generation studies have highlighted the benefits of deep learning-based approaches for generating human-like gestures, these often require large datasets and intensive computation. Our model differentiates between short and long dialogues, generating context-specific...