Few-shot Cross-speaker Style Transfer in Expressive Speech Synthesis