Authors: Wenwen Yu*, Ning Lu*, Xianbiao Qi, Rong Xiao

Affiliation: PAPCIC & XZMU

Description: We propose PICK, a framework that is effective and robust in handling complex documents layout for key information extraction by combining graph learning with graph convolution operation, yielding a richer semantic representation containing the textual and visual features and global layout without ambiguity. For the output of the model, we designed task-specific rules to constrain the final results.