Locate Then Generate: Bridging Vision and Language with Bounding Box for Scene-Text VQA

Re-assign community