Weakly-Supervised Stitching Network for Real-World Panoramic Image Generation

Song, Dae-Young; Lee, Geonsoo; Lee, Hee-Kyung; Um, Gi-Mun; Cho, Donghyeon

doi:10.1007/978-3-031-19787-1_4

Cited by 12 publications

(6 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After that, they proposed ablation constraint to reconstruct the broad scene from feature to pixel (Nie et al 2021). (Song et al 2022) presented a weakly supervised learning method for fisheye panorama generation. Benefiting from the complementary of multi-modality data, (Jiang et al 2022b(Jiang et al , 2023 proposed infrared and visible image stitching.…”

Section: Related Work Image Stitchingmentioning

confidence: 99%

See 1 more Smart Citation

Towards Robust Image Stitching: An Adaptive Resistance Learning against Compatible Attacks

Jiang,

Li,

Liu

et al. 2024

AAAI

View full text Add to dashboard Cite

Image stitching seamlessly integrates images captured from varying perspectives into a single wide field-of-view image. Such integration not only broadens the captured scene but also augments holistic perception in computer vision applications. Given a pair of captured images, subtle perturbations and distortions which go unnoticed by the human visual system tend to attack the correspondence matching, impairing the performance of image stitching algorithms. In light of this challenge, this paper presents the first attempt to improve the robustness of image stitching against adversarial attacks. Specifically, we introduce a stitching-oriented attack (SoA), tailored to amplify the alignment loss within overlapping regions, thereby targeting the feature matching procedure. To establish an attack resistant model, we delve into the robustness of stitching architecture and develop an adaptive adversarial training (AAT) to balance attack resistance with stitching precision. In this way, we relieve the gap between the routine adversarial training and benign models, ensuring resilience without quality compromise. Comprehensive evaluation across real-world and synthetic datasets validate the deterioration of SoA on stitching performance. Furthermore, AAT emerges as a more robust solution against adversarial perturbations, delivering superior stitching results. Code is available at: https://github.com/Jzy2017/TRIS.

show abstract

Section: Related Work Image Stitchingmentioning

confidence: 99%

“…Our base network is built upon PWCnet (Sun et al 2018). This network features a three-scale pyramid designed for effective feature encoding and uses an iterative regression mechanism to achieve correspondence matching in a coarseto-fine manner.…”

Section: Robust Stitching Modelmentioning

confidence: 99%

Towards Robust Image Stitching: An Adaptive Resistance Learning against Compatible Attacks

Jiang,

Li,

Liu

et al. 2024

AAAI

View full text Add to dashboard Cite

show abstract

“…A few recent works have examined image stitching using deep learning [15][16][17][18]. However, all of these focused on pairwise stitching, wherein two inputs are stitched together and the output is stitched with the next input until all inputs have been used, similar to the traditional methods discussed above.…”

Section: Introductionmentioning

confidence: 99%

Semi-Supervised Image Stitching from Unstructured Camera Arrays

Nghonda Tchinda,

Panoff,

Tchuinkou Kwadjo

et al. 2023

Sensors

View full text Add to dashboard Cite

Image stitching involves combining multiple images of the same scene captured from different viewpoints into a single image with an expanded field of view. While this technique has various applications in computer vision, traditional methods rely on the successive stitching of image pairs taken from multiple cameras. While this approach is effective for organized camera arrays, it can pose challenges for unstructured ones, especially when handling scene overlaps. This paper presents a deep learning-based approach for stitching images from large unstructured camera sets covering complex scenes. Our method processes images concurrently by using the SandFall algorithm to transform data from multiple cameras into a reduced fixed array, thereby minimizing data loss. A customized convolutional neural network then processes these data to produce the final image. By stitching images simultaneously, our method avoids the potential cascading errors seen in sequential pairwise stitching while offering improved time efficiency. In addition, we detail an unsupervised training method for the network utilizing metrics from Generative Adversarial Networks supplemented with supervised learning. Our testing revealed that the proposed approach operates in roughly ∼1/7th the time of many traditional methods on both CPU and GPU platforms, achieving results consistent with established methods.

show abstract

“…Image Stitching There are two branches under image stitching, view-fixed [17,22,41,44], and view-free tasks [10,14,21,26,32,33,35,48]. View-fixed scheme stitches images with given fixed views which are free from estimating an aligning transformation.…”

Section: Related Workmentioning

confidence: 99%

“…Estimation of geometric transformations for image stitching requires descriptions of the spatial correspondences between given scenes. To calculate such relationships, existing methods are categorized into two branches: feature-based methods [5,7,8,13,15,16,18,28] and learning-based [10,11,[21][22][23][24][25] methods. Feature-based approaches detect key points or lines to match the textures of images in order to estimate the optimal transformation.…”

Section: Introductionmentioning

confidence: 99%

Republic of Korea (South Korea)

Kim¹

The Crime of Aggression

View full text Add to dashboard Cite

Despite recent progress in semantic image synthesis, complete control over image style remains a challenging problem. Existing methods require reference images to feed style information into semantic layouts, which indicates that the style is constrained by the given image. In this paper, we propose a model named RUCGAN for user controllable semantic image synthesis, which utilizes a singular color to represent the style of a specific semantic region. The proposed network achieves reference-free semantic image synthesis by injecting color as userdesired styles into each semantic layout, and is able to synthesize semantic images with unusual colors. Extensive experimental results on various challenging datasets show that the proposed method outperforms existing methods, and we further provide an interactive UI to demonstrate the advantage of our approach for style controllability. The codes and UI are available at: https://github.com/BenjaminJonghyun/RUCGAN

show abstract

Weakly-Supervised Stitching Network for Real-World Panoramic Image Generation

Cited by 12 publications

References 44 publications

Towards Robust Image Stitching: An Adaptive Resistance Learning against Compatible Attacks

Towards Robust Image Stitching: An Adaptive Resistance Learning against Compatible Attacks

Semi-Supervised Image Stitching from Unstructured Camera Arrays

Republic of Korea (South Korea)

Contact Info

Product

Resources

About