Wei, Xiaokang, Yan, Zizheng, Xiong, Zhangyang, Hao, Yiming, Qin, Yipeng  ORCID: https://orcid.org/0000-0002-1551-9126 and Han, Xiaoguang
2026.
IntrinsicReal: Adapting IntrinsicAnything from synthetic to real objects.
IEEE Transactions on Multimedia
10.1109/TMM.2026.3727723
|
Preview |
PDF
- Accepted Post-Print Version
Download (6MB) | Preview |
Abstract
Estimating albedo from single RGB images captured in real-world environments presents a significant challenge due to the absence of paired ground truth albedos. Therefore, while recent methods (e.g., IntrinsicAnything) have achieved breakthroughs by harnessing powerful diffusion priors, they remain predominantly trained on large-scale synthetic datasets (e.g., Objaverse) and applied directly to real-world RGB images, which ignores the large domain gap between synthetic and real-world data and leads to suboptimal generalization performance. In this work, we address this gap by proposing IntrinsicReal, a novel domain adaptation framework that bridges the above-mentioned domain gap for real-world intrinsic image decomposition. Specifically, IntrinsicReal adapts IntrinsicAnything to the real domain by fine-tuning it using its high-quality output albedos selected by a novel dual pseudo-labeling strategy: i) pseudo-labeling with an absolute confidence threshold on classifier predictions, and ii) pseudo-labeling using the relative preference ranking of classifier predictions for individual input objects. Experimental results show that our IntrinsicReal significantly outperforms existing methods, achieving state-of-the-art results for albedo estimation on both synthetic and real-world datasets.
| Item Type: | Article |
|---|---|
| Date Type: | Published Online |
| Status: | In Press |
| Schools: | Schools > Computational & Mathematical Sciences Schools > Computer Science & Informatics |
| Publisher: | Institute of Electrical and Electronics Engineers |
| ISSN: | 1520-9210 |
| Date of First Compliant Deposit: | 28 August 2026 |
| Date of Acceptance: | 22 June 2026 |
| Last Modified: | 28 Aug 2026 09:45 |
| URI: | https://orca.cardiff.ac.uk/id/eprint/189190 |
Actions (repository staff only)
 |
Edit Item |
Altmetric
Altmetric