A first reproducible quantitative evaluation for automated document layout generation

Deep generative models have been recently experimented in automated document layout generation, which led to significant qualitative results, assessed through user studies and displayed visuals. However, no reproducible quantitative evaluation has been settled in these works, which prevents scientific comparison of upcoming models with previous models. In this context, we propose a turnkey fully reproducible evaluation method for automated document layout generation. This git contains quantitative and visual results, evaluation and data process functions as well as the generated and real layouts on which the evaluation is applied. Results and function calls are displayed and runable from the jupyter notebook.