1 code implementation • 19 Mar 2024 • Hanzhou Liu, Binghan Li, Chengkai Liu, Mi Lu
To this end, we propose DeblurDiNAT, a compact encoder-decoder Transformer which efficiently restores clean images from real-world blurry ones.
Ranked #3 on Deblurring on RealBlur-J (trained on GoPro)
no code implementations • 12 Mar 2021 • Binghan Li, Yindong Hua, Yifeng Liu, Mi Lu
It also reduces the amount of parameters significantly by replacing the fully connected layers with the fully convolutional layers.
no code implementations • 12 Mar 2021 • Binghan Li, Yindong Hua, Mi Lu
To increase object detection accuracy in the hazy environment, the authors further present an algorithm to build a synthetic hazy COCO training dataset by generating the artificial haze to the MS COCO training dataset.
no code implementations • 25 Apr 2019 • Binghan Li, Wenrui Zhang, Mi Lu
The RESIDE dataset provides enough synthetic hazy images and their corresponding groundtruth images to train and test.
no code implementations • 11 Jun 2018 • Ye Wang, Mi Lu
Currently, various types of CAPTCHAs need corresponding segmentation to identify single character due to the numerous different segmentation ways.
no code implementations • 11 Apr 2018 • Han Wang, Ye Wang, Xinxiang Zhang, Mi Lu, Yoonsuck Choe, Jingjing Cao
Unlike previous unknown nouns tagging task, this is the first attempt to focus on out-of-vocabulary (OOV) lexical evaluation tasks that do not require any prior knowledge.