Adapting Pretrained Networks for Image Quality Assessment on High Dynamic Range Displays
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911861469872128 |
|---|---|
| author | Chubarau, Andrei Yoo, Hyunjin Akhavan, Tara Clark, James |
| author_facet | Chubarau, Andrei Yoo, Hyunjin Akhavan, Tara Clark, James |
| contents | Conventional image quality metrics (IQMs), such as PSNR and SSIM, are designed for perceptually uniform gamma-encoded pixel values and cannot be directly applied to perceptually non-uniform linear high-dynamic-range (HDR) colors. Similarly, most of the available datasets consist of standard-dynamic-range (SDR) images collected in standard and possibly uncontrolled viewing conditions. Popular pre-trained neural networks are likewise intended for SDR inputs, restricting their direct application to HDR content. On the other hand, training HDR models from scratch is challenging due to limited available HDR data. In this work, we explore more effective approaches for training deep learning-based models for image quality assessment (IQA) on HDR data. We leverage networks pre-trained on SDR data (source domain) and re-target these models to HDR (target domain) with additional fine-tuning and domain adaptation. We validate our methods on the available HDR IQA datasets, demonstrating that models trained with our combined recipe outperform previous baselines, converge much quicker, and reliably generalize to HDR inputs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_00670 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Adapting Pretrained Networks for Image Quality Assessment on High Dynamic Range Displays Chubarau, Andrei Yoo, Hyunjin Akhavan, Tara Clark, James Computer Vision and Pattern Recognition Image and Video Processing Conventional image quality metrics (IQMs), such as PSNR and SSIM, are designed for perceptually uniform gamma-encoded pixel values and cannot be directly applied to perceptually non-uniform linear high-dynamic-range (HDR) colors. Similarly, most of the available datasets consist of standard-dynamic-range (SDR) images collected in standard and possibly uncontrolled viewing conditions. Popular pre-trained neural networks are likewise intended for SDR inputs, restricting their direct application to HDR content. On the other hand, training HDR models from scratch is challenging due to limited available HDR data. In this work, we explore more effective approaches for training deep learning-based models for image quality assessment (IQA) on HDR data. We leverage networks pre-trained on SDR data (source domain) and re-target these models to HDR (target domain) with additional fine-tuning and domain adaptation. We validate our methods on the available HDR IQA datasets, demonstrating that models trained with our combined recipe outperform previous baselines, converge much quicker, and reliably generalize to HDR inputs. |
| title | Adapting Pretrained Networks for Image Quality Assessment on High Dynamic Range Displays |
| topic | Computer Vision and Pattern Recognition Image and Video Processing |
| url | https://arxiv.org/abs/2405.00670 |