2021
WenLan: Bridging Vision and Language by Large-Scale Multi-Modal Pre-Training.
CoRR, 2021