Adversarial complementary learning for just noticeable difference estimation
Recently, many unsupervised learning-based models have emerged for Just Noticeable Difference (JND) estimation, demonstrating remarkable improvements in accuracy. However, these models suffer from a significant drawback is that their heavy reliance on handcrafted priors for guidance. This restricts...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/178981 |
| _version_ | 1826122120742043648 |
|---|---|
| author | Yu, Dong Jin, Jian Meng, Lili Chen, Zhipeng Zhang, Huaxiang |
| author2 | School of Computer Science and Engineering |
| author_facet | School of Computer Science and Engineering Yu, Dong Jin, Jian Meng, Lili Chen, Zhipeng Zhang, Huaxiang |
| author_sort | Yu, Dong |
| collection | NTU |
| description | Recently, many unsupervised learning-based models have emerged for Just Noticeable Difference (JND) estimation, demonstrating remarkable improvements in accuracy. However, these models suffer from a significant drawback is that their heavy reliance on handcrafted priors for guidance. This restricts the information for estimating JND simply extracted from regions that are highly related to handcrafted priors, while information from the rest of the regions is disregarded, thus limiting the accuracy of JND estimation. To address such issue, on the one hand, we extract the information for estimating JND in an Adversarial Complementary Learning (ACoL) way and propose an ACoL-JND network to estimate the JND by comprehensively considering the handcrafted priors-related regions and non-related regions. On the other hand, to make the handcrafted priors richer, we take two additional priors that are highly related to JND modeling into account, i.e., Patterned Masking (PM) and Contrast Masking (CM). Experimental results demonstrate that our proposed model outperforms the existing JND models and achieves state-of-the-art performance in both subjective viewing tests and objective metrics assessments. |
| first_indexed | 2024-10-01T05:43:17Z |
| format | Journal Article |
| id | ntu-10356/178981 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T05:43:17Z |
| publishDate | 2024 |
| record_format | dspace |
| spelling | ntu-10356/1789812024-07-19T15:35:54Z Adversarial complementary learning for just noticeable difference estimation Yu, Dong Jin, Jian Meng, Lili Chen, Zhipeng Zhang, Huaxiang School of Computer Science and Engineering Computer and Information Science Convolutional neural networks Just noticeable difference Recently, many unsupervised learning-based models have emerged for Just Noticeable Difference (JND) estimation, demonstrating remarkable improvements in accuracy. However, these models suffer from a significant drawback is that their heavy reliance on handcrafted priors for guidance. This restricts the information for estimating JND simply extracted from regions that are highly related to handcrafted priors, while information from the rest of the regions is disregarded, thus limiting the accuracy of JND estimation. To address such issue, on the one hand, we extract the information for estimating JND in an Adversarial Complementary Learning (ACoL) way and propose an ACoL-JND network to estimate the JND by comprehensively considering the handcrafted priors-related regions and non-related regions. On the other hand, to make the handcrafted priors richer, we take two additional priors that are highly related to JND modeling into account, i.e., Patterned Masking (PM) and Contrast Masking (CM). Experimental results demonstrate that our proposed model outperforms the existing JND models and achieves state-of-the-art performance in both subjective viewing tests and objective metrics assessments. Published version This work was supported in part by the NSF of Shandong Province under Grant ZR2020MF042 and Grant ZR2022MF346; Science and Technology Plan Project of Tangshan Science and Technology Bureau Tangshan Foundation Innovation Team of Digital Media Security under Grant 21130212D. 2024-07-15T05:52:38Z 2024-07-15T05:52:38Z 2024 Journal Article Yu, D., Jin, J., Meng, L., Chen, Z. & Zhang, H. (2024). Adversarial complementary learning for just noticeable difference estimation. KSII Transactions On Internet and Information Systems, 18(2), 438-455. https://dx.doi.org/10.3837/TIIS.2024.02.009 1976-7277 https://hdl.handle.net/10356/178981 10.3837/TIIS.2024.02.009 2-s2.0-85193392191 2 18 438 455 en KSII Transactions on Internet and Information Systems © 2024 KSII. This is an open-access article distributed under the terms of the Creative Commons License. application/pdf |
| spellingShingle | Computer and Information Science Convolutional neural networks Just noticeable difference Yu, Dong Jin, Jian Meng, Lili Chen, Zhipeng Zhang, Huaxiang Adversarial complementary learning for just noticeable difference estimation |
| title | Adversarial complementary learning for just noticeable difference estimation |
| title_full | Adversarial complementary learning for just noticeable difference estimation |
| title_fullStr | Adversarial complementary learning for just noticeable difference estimation |
| title_full_unstemmed | Adversarial complementary learning for just noticeable difference estimation |
| title_short | Adversarial complementary learning for just noticeable difference estimation |
| title_sort | adversarial complementary learning for just noticeable difference estimation |
| topic | Computer and Information Science Convolutional neural networks Just noticeable difference |
| url | https://hdl.handle.net/10356/178981 |
| work_keys_str_mv | AT yudong adversarialcomplementarylearningforjustnoticeabledifferenceestimation AT jinjian adversarialcomplementarylearningforjustnoticeabledifferenceestimation AT menglili adversarialcomplementarylearningforjustnoticeabledifferenceestimation AT chenzhipeng adversarialcomplementarylearningforjustnoticeabledifferenceestimation AT zhanghuaxiang adversarialcomplementarylearningforjustnoticeabledifferenceestimation |