Describe Anything Model for Visual Question Answering on Text-rich Images Paper • 2507.12441 • Published Jul 16, 2025 • 5