Deep HDR Hallucination for Inverse Tone Mapping
Inverse Tone Mapping (ITM) methods attempt to reconstruct High Dynamic Range (HDR) information from Low Dynamic Range (LDR) image content. The dynamic range of well-exposed areas must be expanded and any missing information due to over/under-exposure must be recovered (hallucinated). The majority of...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/12/4032 |
| _version_ | 1797530524779020288 |
|---|---|
| author | Demetris Marnerides Thomas Bashford-Rogers Kurt Debattista |
| author_facet | Demetris Marnerides Thomas Bashford-Rogers Kurt Debattista |
| author_sort | Demetris Marnerides |
| collection | DOAJ |
| description | Inverse Tone Mapping (ITM) methods attempt to reconstruct High Dynamic Range (HDR) information from Low Dynamic Range (LDR) image content. The dynamic range of well-exposed areas must be expanded and any missing information due to over/under-exposure must be recovered (hallucinated). The majority of methods focus on the former and are relatively successful, while most attempts on the latter are not of sufficient quality, even ones based on Convolutional Neural Networks (CNNs). A major factor for the reduced inpainting quality in some works is the choice of loss function. Work based on Generative Adversarial Networks (GANs) shows promising results for image synthesis and LDR inpainting, suggesting that GAN losses can improve inverse tone mapping results. This work presents a GAN-based method that hallucinates missing information from badly exposed areas in LDR images and compares its efficacy with alternative variations. The proposed method is quantitatively competitive with state-of-the-art inverse tone mapping methods, providing good dynamic range expansion for well-exposed areas and plausible hallucinations for saturated and under-exposed areas. A density-based normalisation method, targeted for HDR content, is also proposed, as well as an HDR data augmentation method targeted for HDR hallucination. |
| first_indexed | 2024-03-10T10:30:13Z |
| format | Article |
| id | doaj.art-5653262404ff4b4eb20feb092817c702 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T10:30:13Z |
| publishDate | 2021-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-5653262404ff4b4eb20feb092817c7022023-11-21T23:42:34ZengMDPI AGSensors1424-82202021-06-012112403210.3390/s21124032Deep HDR Hallucination for Inverse Tone MappingDemetris Marnerides0Thomas Bashford-Rogers1Kurt Debattista2WMG, University of Warwick, Coventry CV4 7AL, UKDepartment of Computer Science and Creative Technologies, University of the West of England, Bristol BS16 1GY, UKWMG, University of Warwick, Coventry CV4 7AL, UKInverse Tone Mapping (ITM) methods attempt to reconstruct High Dynamic Range (HDR) information from Low Dynamic Range (LDR) image content. The dynamic range of well-exposed areas must be expanded and any missing information due to over/under-exposure must be recovered (hallucinated). The majority of methods focus on the former and are relatively successful, while most attempts on the latter are not of sufficient quality, even ones based on Convolutional Neural Networks (CNNs). A major factor for the reduced inpainting quality in some works is the choice of loss function. Work based on Generative Adversarial Networks (GANs) shows promising results for image synthesis and LDR inpainting, suggesting that GAN losses can improve inverse tone mapping results. This work presents a GAN-based method that hallucinates missing information from badly exposed areas in LDR images and compares its efficacy with alternative variations. The proposed method is quantitatively competitive with state-of-the-art inverse tone mapping methods, providing good dynamic range expansion for well-exposed areas and plausible hallucinations for saturated and under-exposed areas. A density-based normalisation method, targeted for HDR content, is also proposed, as well as an HDR data augmentation method targeted for HDR hallucination.https://www.mdpi.com/1424-8220/21/12/4032high dynamic rangeinverse tone mappingdeep learning |
| spellingShingle | Demetris Marnerides Thomas Bashford-Rogers Kurt Debattista Deep HDR Hallucination for Inverse Tone Mapping Sensors high dynamic range inverse tone mapping deep learning |
| title | Deep HDR Hallucination for Inverse Tone Mapping |
| title_full | Deep HDR Hallucination for Inverse Tone Mapping |
| title_fullStr | Deep HDR Hallucination for Inverse Tone Mapping |
| title_full_unstemmed | Deep HDR Hallucination for Inverse Tone Mapping |
| title_short | Deep HDR Hallucination for Inverse Tone Mapping |
| title_sort | deep hdr hallucination for inverse tone mapping |
| topic | high dynamic range inverse tone mapping deep learning |
| url | https://www.mdpi.com/1424-8220/21/12/4032 |
| work_keys_str_mv | AT demetrismarnerides deephdrhallucinationforinversetonemapping AT thomasbashfordrogers deephdrhallucinationforinversetonemapping AT kurtdebattista deephdrhallucinationforinversetonemapping |