MMA: Multi-Modal Adapter for Vision-Language Models | IEEE Conference Publication | IEEE Xplore