Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.09455 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911101875126272 |
|---|---|
| author | Safwan, Hussain Md. Mahim, Mahbub Islam |
| author_facet | Safwan, Hussain Md. Mahim, Mahbub Islam |
| contents | A common dilemma while photographing a scene is whether to capture it at a wider angle, allowing more of the scene to be covered but in less detail or to click in a narrow angle that captures better details but leaves out portions of the scene. We propose a novel method in this paper that infuses wider shots with finer quality details that is usually associated with an image captured by the primary lens by capturing the same scene using both narrow and wide field of view (FoV) lenses. We do so by training a Generative Adversarial Network (GAN)-based model to learn to extract the visual quality parameters from a narrow-angle shot and to transfer these to the corresponding wide-angle image of the scene using residual connections and an attention-based fusion module. We have mentioned in details the proposed technique to isolate the visual essence of an image and to transfer it into another image. We have also elaborately discussed our implementation details and have presented the results of evaluation over several benchmark datasets and comparisons with contemporary advancements in the field. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_09455 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Enhancing Wide-Angle Image Using Narrow-Angle View of the Same Scene Safwan, Hussain Md. Mahim, Mahbub Islam Computer Vision and Pattern Recognition Image and Video Processing F.2.2; I.2.7 A common dilemma while photographing a scene is whether to capture it at a wider angle, allowing more of the scene to be covered but in less detail or to click in a narrow angle that captures better details but leaves out portions of the scene. We propose a novel method in this paper that infuses wider shots with finer quality details that is usually associated with an image captured by the primary lens by capturing the same scene using both narrow and wide field of view (FoV) lenses. We do so by training a Generative Adversarial Network (GAN)-based model to learn to extract the visual quality parameters from a narrow-angle shot and to transfer these to the corresponding wide-angle image of the scene using residual connections and an attention-based fusion module. We have mentioned in details the proposed technique to isolate the visual essence of an image and to transfer it into another image. We have also elaborately discussed our implementation details and have presented the results of evaluation over several benchmark datasets and comparisons with contemporary advancements in the field. |
| title | Enhancing Wide-Angle Image Using Narrow-Angle View of the Same Scene |
| topic | Computer Vision and Pattern Recognition Image and Video Processing F.2.2; I.2.7 |
| url | https://arxiv.org/abs/2504.09455 |