ViSMoE: Visual-Aware Sparse Mixture-of-Experts for Embodied Referring Expression Grounding.

Shuo Feng, Piji Li

Browse the full ICANN paper archive.