Saved in:
Bibliographic Details
Main Authors: Garg, Aakash, Zeng, Libing, Tsarov, Andrii, Kalantari, Nima Khademi
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2506.05367
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909700907335680
author Garg, Aakash
Zeng, Libing
Tsarov, Andrii
Kalantari, Nima Khademi
author_facet Garg, Aakash
Zeng, Libing
Tsarov, Andrii
Kalantari, Nima Khademi
contents In this paper, we propose a novel diffusion-based approach to generate stereo images given a text prompt. Since stereo image datasets with large baselines are scarce, training a diffusion model from scratch is not feasible. Therefore, we propose leveraging the strong priors learned by Stable Diffusion and fine-tuning it on stereo image datasets to adapt it to the task of stereo generation. To improve stereo consistency and text-to-image alignment, we further tune the model using prompt alignment and our proposed stereo consistency reward functions. Comprehensive experiments demonstrate the superiority of our approach in generating high-quality stereo images across diverse scenarios, outperforming existing methods.
format Preprint
id arxiv_https___arxiv_org_abs_2506_05367
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Text2Stereo: Repurposing Stable Diffusion for Stereo Generation with Consistency Rewards
Garg, Aakash
Zeng, Libing
Tsarov, Andrii
Kalantari, Nima Khademi
Computer Vision and Pattern Recognition
In this paper, we propose a novel diffusion-based approach to generate stereo images given a text prompt. Since stereo image datasets with large baselines are scarce, training a diffusion model from scratch is not feasible. Therefore, we propose leveraging the strong priors learned by Stable Diffusion and fine-tuning it on stereo image datasets to adapt it to the task of stereo generation. To improve stereo consistency and text-to-image alignment, we further tune the model using prompt alignment and our proposed stereo consistency reward functions. Comprehensive experiments demonstrate the superiority of our approach in generating high-quality stereo images across diverse scenarios, outperforming existing methods.
title Text2Stereo: Repurposing Stable Diffusion for Stereo Generation with Consistency Rewards
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2506.05367