Graduate Student(Shanxi University 2020-2023)
Computer Vision Algorithm Engineer(2023-cur Beijing QiHu 360 AI Institute)
My main direction is low-level vision tasks such as super-resolution, image dehazing, low-light enhancement and image derain.
My research interests focus on unified enhancement models for multiple classes of low-quality images. Strategies such as incremental learning, multimodal guidance, contrastive learning, self-supervised learning, and multi-stage inpainting are being attempted to tackle this challenge.
Use multi-modal, large model and other strategies to achieve open-world obeject detection, unified weather repair large model.
Self-supervised Learning and Adaptation for Single Image Dehazing(IJCAI-ECAI2022 Long Presentation)
link:(https://github.com/DongLiangSXU/SLAdehazing)
Learning Hierarchical Dynamics with Spatial Adjacency for Image Enhancement.(ACMMM2022 Poster)