Exploring Self-Supervised Speech Representations for Cross-lingual Acoustic-to-Articulatory Inversion | Synapse