StereoINR: Cross-View Geometry Consistent Stereo Super Resolution with Implicit Neural Representation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Yi, Liu, Xinyi, Wan, Yi, Xia, Panwang, Wu, Qiong, Zhang, Yongjun
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911037837541376
author Liu, Yi
Liu, Xinyi
Wan, Yi
Xia, Panwang
Wu, Qiong
Zhang, Yongjun
author_facet Liu, Yi
Liu, Xinyi
Wan, Yi
Xia, Panwang
Wu, Qiong
Zhang, Yongjun
contents Stereo image super-resolution (SSR) aims to enhance high-resolution details by leveraging information from stereo image pairs. However, existing stereo super-resolution (SSR) upsampling methods (e.g., pixel shuffle) often overlook cross-view geometric consistency and are limited to fixed-scale upsampling. The key issue is that previous upsampling methods use convolution to independently process deep features of different views, lacking cross-view and non-local information perception, making it difficult to select beneficial information from multi-view scenes adaptively. In this work, we propose Stereo Implicit Neural Representation (StereoINR), which innovatively models stereo image pairs as continuous implicit representations. This continuous representation breaks through the scale limitations, providing a unified solution for arbitrary-scale stereo super-resolution reconstruction of left-right views. Furthermore, by incorporating spatial warping and cross-attention mechanisms, StereoINR enables effective cross-view information fusion and achieves significant improvements in pixel-level geometric consistency. Extensive experiments across multiple datasets show that StereoINR outperforms out-of-training-distribution scale upsampling and matches state-of-the-art SSR methods within training-distribution scales.
format Preprint
id arxiv_https___arxiv_org_abs_2505_05509
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle StereoINR: Cross-View Geometry Consistent Stereo Super Resolution with Implicit Neural Representation
Liu, Yi
Liu, Xinyi
Wan, Yi
Xia, Panwang
Wu, Qiong
Zhang, Yongjun
Image and Video Processing
Computer Vision and Pattern Recognition
Stereo image super-resolution (SSR) aims to enhance high-resolution details by leveraging information from stereo image pairs. However, existing stereo super-resolution (SSR) upsampling methods (e.g., pixel shuffle) often overlook cross-view geometric consistency and are limited to fixed-scale upsampling. The key issue is that previous upsampling methods use convolution to independently process deep features of different views, lacking cross-view and non-local information perception, making it difficult to select beneficial information from multi-view scenes adaptively. In this work, we propose Stereo Implicit Neural Representation (StereoINR), which innovatively models stereo image pairs as continuous implicit representations. This continuous representation breaks through the scale limitations, providing a unified solution for arbitrary-scale stereo super-resolution reconstruction of left-right views. Furthermore, by incorporating spatial warping and cross-attention mechanisms, StereoINR enables effective cross-view information fusion and achieves significant improvements in pixel-level geometric consistency. Extensive experiments across multiple datasets show that StereoINR outperforms out-of-training-distribution scale upsampling and matches state-of-the-art SSR methods within training-distribution scales.
title StereoINR: Cross-View Geometry Consistent Stereo Super Resolution with Implicit Neural Representation
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2505.05509