VLM-CD: Making Concept Annotations Optional with Frozen Vision-Language Embeddings
Submitted to AAAI 2027 on July 20, 2026
VLM-CD explores frozen vision-language representations as a way to reduce the need for manually supplied concept annotations in cognitive diagnosis.
Recommended citation: Wang, Z., Jiang, H., Hou, M., Bai, Y., Luo, R., Guo, T., Zheng, J., & Liang, Q. (2027). "VLM-CD: Making Concept Annotations Optional with Frozen Vision-Language Embeddings." AAAI 2027. Conference submission.