Vision language models have difficulty recognizing virtual objects

Re-assign community