On Efficient Transformer-Based Image Pre-training for Low-Level Vision

Easy-to-use,Modular and Extendible package of deep-learning based CTR models .

Simple realization of papers in oppo Research Institute super score competition.

[CVPR 2021] This is the official code for the paper "Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement"

Unofficial implementation of "TableNet: Deep Learning model for end-to-end Table detection and Tabular data extraction from Scanned Document Images"

ctcloss + centerloss crnn text recognition

Document photo perspective transformation correction，利用unet网络进行 文档照片透视纠偏