no code implementations • 16 Apr 2024 • Kang Liao, Zongsheng Yue, Zhonghua Wu, Chen Change Loy
To our knowledge, this is the first work that solves multiple practical warping tasks in one single model.
no code implementations • 15 Apr 2024 • Jiyuan Wang, Chunyu Lin, Lang Nie, Kang Liao, Shuwei Shao, Yao Zhao
In this paper, we propose a novel robust depth estimation method called D4RD, featuring a custom contrastive learning mode tailored for diffusion models to mitigate performance degradation in complex environments.
1 code implementation • 11 Mar 2024 • Lang Nie, Chunyu Lin, Kang Liao, Yun Zhang, Shuaicheng Liu, Yao Zhao
In this paper, we retarget video stitching to an emerging issue, named warping shake, when extending image stitching to video stitching.
1 code implementation • 24 Jan 2024 • Lang Nie, Chunyu Lin, Kang Liao, Shuaicheng Liu, Yao Zhao
To break this bottleneck, we propose the coupled thin-plate spline model (CoupledTPS), which iteratively couples multiple TPS with limited control points into a more flexible and powerful transformation.
no code implementations • 26 Dec 2023 • Zhijie Shen, Chunyu Lin, Junsong Zhang, Lang Nie, Kang Liao, Yao Zhao
Existing panoramic layout estimation solutions tend to recover room boundaries from a vertically compressed sequence, yielding imprecise results as the compression process often muddles the semantics between various planes.
1 code implementation • 19 Mar 2023 • Kang Liao, Lang Nie, Shujuan Huang, Chunyu Lin, Jing Zhang, Yao Zhao, Moncef Gabbouj, DaCheng Tao
In this paper, we provide a comprehensive survey of learning-based camera calibration techniques, by analyzing their strengths and limitations.
1 code implementation • CVPR 2023 • Zhijie Shen, Zishuo Zheng, Chunyu Lin, Lang Nie, Kang Liao, Shuai Zheng, Yao Zhao
Based on the Manhattan World assumption, most existing indoor layout estimation schemes focus on recovering layouts from vertically compressed 1D sequences.
no code implementations • 20 Feb 2023 • Zisong Chen, Chunyu Lin, Lang Nie, Kang Liao, Yao Zhao
In this paper, we propose the first unsupervised omnidirectional MVS framework based on multiple fisheye images.
1 code implementation • ICCV 2023 • Lang Nie, Chunyu Lin, Kang Liao, Shuaicheng Liu, Yao Zhao
First, we propose a robust and flexible warp to model the image registration from global homography to local thin-plate spline motion.
1 code implementation • 8 Feb 2023 • Shangrong Yang, Chunyu Lin, Kang Liao, Yao Zhao
Subsequently, we observe that the inter-frame optical flow of the video is facilitated to perceive the local spatial deformation of the fisheye video.
no code implementations • 26 Jan 2023 • Shangrong Yang, Chunyu Lin, Kang Liao, Yao Zhao
To this end, we propose a Dual Diffusion Architecture (DDA) for the fisheye rectification with a better generalization ability.
1 code implementation • ICCV 2023 • Kang Liao, Lang Nie, Chunyu Lin, Zishuo Zheng, Yao Zhao
In this work, we explore constructing a win-win representation on both content and boundary by contributing a new learning model, i. e., Rectangling Rectification Network (RecRecNet).
no code implementations • ICCV 2023 • Shangrong Yang, Chunyu Lin, Kang Liao, Yao Zhao
Fisheye image rectification is hindered by synthetic models producing poor results for real-world correction.
1 code implementation • 3 Aug 2022 • Zhijie Shen, Chunyu Lin, Lang Nie, Kang Liao, Yao Zhao
For a monocular 360 image, depth estimation is a challenging because the distortion increases along the latitude.
Ranked #8 on Depth Estimation on Stanford2D3D Panoramic
1 code implementation • 7 Jul 2022 • Lang Nie, Chunyu Lin, Kang Liao, Shuaicheng Liu, Yao Zhao
To this end, we leverage a neural network to predict the optical flows that can warp the tilted images to be perceptually horizontal.
no code implementations • 6 Jul 2022 • Zishuo Zheng, Chunyu Lin, Lang Nie, Kang Liao, Zhijie Shen, Yao Zhao
In this paper, we combine the two different representations and propose a novel 360{\deg} semantic segmentation solution from a complementary perspective.
no code implementations • 5 Jul 2022 • Shangrong Yang, Chunyu Lin, Kang Liao, Yao Zhao
To leverage these two characteristics, we introduced Fishformer that processes the fisheye image as a sequence to enhance global and local perception.
1 code implementation • 12 Jun 2022 • Kang Liao, Chunyu Lin, Yunchao Wei, Yao Zhao
For the distortion synthesis, we propose a spiral distortion-aware perception module, in which the learning path keeps consistent with the distortion prior of the fisheye image.
1 code implementation • 18 Apr 2022 • Kang Liao, Xiangyu Xu, Chunyu Lin, Wenqi Ren, Yunchao Wei, Yao Zhao
Motivated by this analysis, we present a Cylin-Painting framework that involves meaningful collaborations between inpainting and outpainting and efficiently fuses the different arrangements, with a view to leveraging their complementary benefits on a seamless cylinder.
no code implementations • 12 Apr 2022 • Lei Zhang, Kang Liao, Chunyu Lin, Yao Zhao
Concretely, we propose a Depth-Guided Outpainting Network to model different feature representations of two modalities and learn the structure-aware cross-modal fusion.
no code implementations • 18 Mar 2022 • Zhijie Shen, Chunyu Lin, Lang Nie, Kang Liao, Yao Zhao
It comprises two modules: Dual-Cubemap Depth Estimation (DCDE) module and Boundary Revision (BR) module.
1 code implementation • 17 Mar 2022 • Zhijie Shen, Chunyu Lin, Kang Liao, Lang Nie, Zishuo Zheng, Yao Zhao
In particular, we divide patches on the spherical tangent domain into tokens to reduce the negative effect of panoramic distortions.
Ranked #4 on Depth Estimation on Stanford2D3D Panoramic
1 code implementation • CVPR 2022 • Lang Nie, Chunyu Lin, Kang Liao, Shuaicheng Liu, Yao Zhao
In this paper, we address these issues by proposing the first deep learning solution to image rectangling.
1 code implementation • 6 Jul 2021 • Lang Nie, Chunyu Lin, Kang Liao, Shuaicheng Liu, Yao Zhao
Homography estimation is an important task in computer vision applications, such as image stitching, video stabilization, and camera calibration.
1 code implementation • 24 Jun 2021 • Lang Nie, Chunyu Lin, Kang Liao, Shuaicheng Liu, Yao Zhao
Even compared with the supervised solutions, our image stitching quality is still preferred by users.
1 code implementation • CVPR 2021 • Shangrong Yang, Chunyu Lin, Kang Liao, Chunjie Zhang, Yao Zhao
We embed a correction layer in skip-connection and leverage the appearance flows in different layers to pre-correct the image features.
no code implementations • ICCV 2021 • Kang Liao, Chunyu Lin, Yunchao Wei, Feng Li, Shangrong Yang, Yao Zhao
To our knowledge, we are the first to tackle the challenging rectification via outpainting, and our curve-aware strategy can reach a rectification construction with complete content and regular shape.
no code implementations • ICCV 2021 • Kang Liao, Chunyu Lin, Lixin Liao, Yao Zhao, Weiyao Lin
In this paper, inspired by the curriculum learning, we analyze the barrel distortion rectification task in a progressive and meaningful manner.
no code implementations • 11 Dec 2020 • Lang Nie, Chunyu Lin, Kang Liao, Yao Zhao
In this paper, we propose an image stitching learning framework, which consists of a large-baseline deep homography module and an edge-preserved deformation module.
1 code implementation • 21 Jul 2020 • Kang Liao, Chunyu Lin, Yao Zhao
Our key insight is that distortion rectification can be cast as a problem of learning an ordinal distortion from a single distorted image.
no code implementations • 20 Jun 2020 • Haojie Liu, Kang Liao, Chunyu Lin, Yao Zhao, Yulan Guo
Pseudo-LiDAR point cloud interpolation is a novel and challenging task in the field of autonomous driving, which aims to address the frequency mismatching problem between camera and LiDAR.
no code implementations • 16 Sep 2019 • Haojie Liu, Kang Liao, Chunyu Lin, Yao Zhao, Yulan Guo
In this paper, we propose a novel Pseudo-LiDAR interpolation network (PLIN) to increase the frequency of LiDAR sensors.