Enhancing Multimodal Information Extraction from Visually Rich Documents with 2D Positional Embeddings.

Aresha Arshad, Momina Moetesum, Adnan Ul-Hasan, Faisal Shafait

Browse the full DICTA paper archive.