MoE-ViE: Mixture of Experts Vision Encoder for Efficient Image and Video Understanding

arXiv:2608.17402v1 Announce Type: new Abstract: Vision encoders are a critical component of vision-language models, and scaling their capacity effectively improves performance. However, dense scaling increases compute cost and inference latency. Mixture-of-Experts (MoE) architectures offer a compelling alternative, having enabled efficient scaling in LLMs, yet the MoE design space for CLIP-style vision encoders remains underexplored at State-of-the-Art (SOTA) levels. In this work, we systematica...