VLMixer: Unpaired Vision-Language Pre-training via Cross-Modal CutMix

Publication
International Conference on Machine Learning (ICML) 2022

Related