Do Vision-Language Foundational models show Robust Visual Perception?