In this paper, we review the NTIRE 2026 challenge on single-image reflection removal (SIRR) in the wild. SIRR is a fundamental task in image restoration. Despite progress in academic research, most methods are tested on synthetic images or limited real-world images, creating a gap in real-world applications. In this challenge, we provide participants with the OpenRR-5k dataset. This dataset requires participants to process real-world images covering a range of reflection scenarios and intensities, aiming to generate clean images without reflections. The challenge attracted more than 100 registrations, with eleven of them participating in the final testing phase. The top-ranked methods advanced the state-of-the-art reflection removal performance and earned unanimous recognition from five experts in the field. The proposed OpenRR-5k dataset is available at https://huggingface.co/datasets/qiuzhangTiTi/OpenRR-5k, and the homepage of this challenge is at https://github.com/caijie0620/OpenRR-5k.
Figures & tables
Methods
Venue
Scheme
Cross-stage fusion
Single-stage
Zhang et al. [ 81 ]
CVPR 2018
I→[T,R]
N/A
ERRNet [ 70 ]
CVPR 2019
I→T
N/A
RobustSIRR [ 56 ]
CVPR 2023
Imultiscale→T
N/A
YTMT [ 26 ]
NeurIPS 2021
I→[T,R]
N/A
F2T2-HiT [ 6 ]
ICIP 2025
I→T
N/A
Two-stage
CoRRN [ 64 ]
TPAMI 2019
I→ET[I,ET]→T
Convolutional Fusion
Table 1: I , R , T , and E represent the I nput, R eflection, T ransmission, and E dge map, respectively. The subscripts of T and R represent intermediate process outputs. The Absorption Effect e is introduced in [ 84 ] to describe light attenuation as it passes through the glass. The output residue term, proposed in [ 27 ] , is used to correct errors in the additive reconstruction of the reflection and transmission layers. Language descriptions in [ 85 ] provide contextual information about the image layers, assisting in addressing the ill-posed nature of the reflection separation problem.
Figure 1: Visualization of paired data generation pipeline for reflection removal.
Figure 2: Visualization of the OPPO AI Reflection Remover pipeline deployed on Find X8 Ultra, using the April 2025 algorithm version. The OPPO reflection removal method adopts a two-stage framework: (1) In the first stage, the system utilizes YOLO and YOSO for robust detection of strong reflection masks, followed by LaMa-based inpainting to fill the detected regions. (2) In the second stage, NAFNet is employed to perform fine-grained reflection removal, further restoring the underlying transmission layer and ensuring high visual fidelity.
Figure 3: Samples of the OpenRR-5k dataset.
Figure 4: The category distribution of our OpenRR-5k test dataset.
Figure 5: Visual comparison on test0004 shows that top methods effectively remove reflections while preserving details, whereas lower-ranked methods suffer from residual reflections, over-smoothing, or color distortions, consistent with their lower subjective scores. For instance, while methods from VIP Lab, YuFans, and KLETech-CEVI are capable of suppressing intense reflections, they struggle to faithfully recover the underlying background details.
Figure 6: Visual comparison across diverse OpenRR-5k test real-world scenes. Top-ranked methods (RRay, Xreflect Master, and AIIALab) consistently achieve superior reflection removal while preserving fine details and faithful colors. In contrast, lower-ranked methods often exhibit residual reflections, over-smoothing, or color distortions, which aligns with their lower subjective rankings. The ten rows from top to bottom correspond to test samples test0005 , test0006 , test0023 , test0026 , test0029 , test0049 , test0075 , test0077 , test0096 , and test0099 , respectively.
Validation Set
Test Set
Team Name
PSNR ↑
SSIM ↑
LPIPS ↓
DISTS ↓
NIQE ↓
Subjective Score ↑
RRay
36.1688 (1)
0.9758 (3)
0.0235 (4)
0.0135 (2)
3.7375 (2)
4.45 (1)
Xreflect Master
36.0496 (2)
0.9776 (1)
0.0210 (1)
0.0127 (1)
3.7648 (6)
4.31 (2)
AIIALab
35.3799 (3)
0.9750 (4)
0.0231 (2)
0.0155 (5)
3.7737 (9)
4.23 (3)
VIP Lab
34.6872 (5)
0.9766 (2)
0.0231 (2)
0.0148 (3)
3.7218 (1)
3.85 (4)
YuFans
34.9062 (4)
0.9738 (8)
0.0257 (6)
0.0159 (6)
3.7783 (10)
3.59 (5)
Table 2: Objective and subjective results on the NTIRE 2026 SIRR Challenge. Note: objective metrics are computed on the OpenRR-5k val ; subjective scores are calculated on the OpenRR-5k test . NTIRE 2026 SIRR in the Wild Challenge Award Winners: 1st Prize: RRay; 2nd Prizes: Xreflect Master, AIIALab; 3rd Prizes: VIP Lab, YuFans, KLETech-CEVI.
Figure 7: An overview of RdNafNet for SIRR.
Figure 8: Overview of the proposed RDNet-XL framework with diffusion-based distillation.
Figure 9: An overview of MS-RDNet for SIRR. Code Link: https://github.com/BrewieRain/MS-RDNet
Figure 10: The overall architecture of the proposed network. Code Link: github.com/j0nghyukpark/CompSIRS
Figure 11: Overview of our approach. Top: Two-stage fine-tuning pipeline with SWA and geometric TTA. Numbers on arrows indicate validation PSNR at each stage. Bottom: RDNet architecture with FocalNet-Large backbone, RevCol body, and three separate NAFBlock decoders. Code Link: Google Drive
Figure 12: Overview of the proposed reflection removal framework based on the XReflection architecture. The network learns to recover the clean transmission layer from a reflection-contaminated input image through hierarchical feature extraction and residual reflection estimation. Code Link: https://github.com/CeviKle/NTIRE2026-KLETech-CEVI-SIRR.git
Figure 13: DUSKAN architecture. Symmetric 4-level U-Net with DUSKANBlock stages. Path A (blue) extracts global features via FFT magnitude modulation, enriched with spectral positional encoding and SE reweighting. Path B (red) uses Kolmogorov-Arnold polynomial-basis activations with a parallel-additive selective gate. A learned per-stage logit α blends both outputs. Code Link: Google Drive
Figure 14: Overview of the final SiGMoid pipeline used for the NTIRE 2026 SIRR final submission. Code Link: Google Drive
Figure 15: TimeDiffiT architecture. The doubly-corrupted input X~tM and noise level σt enter the time-conditioned U-Net (encoder: orange; decoder: blue), producing the restored output X^ . During SFT, no masking is applied and σt=0 .
Figure 16: Overview of our proposed SIRR-Net for single image reflection removal. The U-Net encoder-decoder processes the blended input through four resolution levels, with each stage comprising stacked Transformer Blocks (MDTA + GDFN + SE). Skip connections fuse encoder and decoder features. A global residual connection adds the input to the output. Code Link: https://github.com/Shreenikethjoshi/NTIRE2026-Reflection-removal-in-the-wild
Figure 17: An overview of RDNet+ for SIRR. Code Link: https://github.com/wuw2135/RDNet-with-Frozen-DINOv2-Semantic-Prior-for-Reflection-Removal
Figure 18: Visual comparison of reflection removal results. From left to right: input images, and results generated by OPPO-baseline, RRay, gpt-image-1.5, gpt-image-2, Nano Banana Pro, and Nano Banana 2. The generative methods (gpt-image series and Nano Banana series) demonstrate superior performance in restoring background details and suppressing artifacts. However, as seen in the second row, generative models may occasionally misidentify intense reflections as background structures, leading to unintended artifact amplification.
Prior dual-stream methods with the feature interaction mechanism have achieved remarkable performance in single image reflection removal (SIRR). However, they often struggle with (1) semantic understanding gap between the features of pre-trained models and those of reflection removal models, and (2) reflection label inconsistencies between synthetic and real-world training data. In this work, we first adopt the parameter efficient fine-tuning (PEFT) strategy by integrating several learnable Mona layers into the pre-trained model to align the training directions. Then, a label generator is designed to unify the reflection labels for both synthetic and real-world data. In addition, a Gaussian-based Adaptive Frequency Learning Block (G-AFLB) is proposed to adaptively learn and fuse the frequency priors, and a Dynamic Agent Attention (DAA) is employed as an alternative to window-based attention by dynamically modeling the significance levels across windows (inter-) and within an individual window (intra-). These components constitute our proposed Gap-Free Reflection Removal Network (GFRRN). Extensive experiments demonstrate the effectiveness of our GFRRN, achieving superior performance against state-of-the-art SIRR methods.
Yu Chen, Zewei He, Xingyu Liu +2
Zhejiang University · Huanjiang Laboratory · The Chinese University of Hong Kong
Single-image reflection removal aims to recover a clean transmission layer from one image captured through glass. We study an explicit decomposition pipeline built on RDNet and introduce LowAux, a training-only low-pass reflection auxiliary objective. The original residual target remains the main reflection supervision, while symmetrically filtered prediction and target provide a stable low-frequency constraint. We further incorporate scene-balanced real pairs from RRW to broaden real-scene coverage and improve cross-dataset generalization. To avoid evaluation discrepancies caused by model-specific resizing, padding, output quantization, and metric code, we build a unified public benchmark over CEILNet, Real20, Postcard, Objects, and Wild. Under the same evaluator, the proposed system obtains a five-dataset macro average of 27.546 dB PSNR, 0.9220 SSIM, 0.9751 NCC, and 0.004760 LMSE, achieving the highest macro-average PSNR, SSIM, and NCC and the lowest LMSE among the compared public checkpoints and internal variants. Per-dataset and qualitative analyses show that the main benefit is a more balanced performance across diverse reflection distributions, while clear semantic reflections in Postcard remain challenging.
Single-image reflection removal (SIRR) aims to recover the clean transmission layer from a reflection-contaminated image. Although recent methods achieve promising results with large diffusion models, they rely on image-agnostic adaptation strategies, e.g., fine-tuning or ControlNet, that enforce uniform suppression regardless of reflection severity. As a result, heavy reflections often leave residuals, while weak ones suffer from detail loss. To this end, we propose ReLo-IRR, a reflection-guided LoRA framework built upon the rectified flow model. First, a lightweight estimator is designed to predict the reflection strength descriptor, providing an explicit prior of reflection dominance for each image and enabling image-dependent LoRA modulation. Second, we introduce a time-conditioned mechanism that fuses this reflection descriptor with timestep embeddings, enabling LoRA modulation to evolve consistently with the coarse-to-fine denoising process. By jointly modeling reflection strength and denoising dynamics, our ReLo-IRR achieves robust suppression of diverse reflection conditions. Extensive experiments on challenging benchmarks validate the effectiveness of ReLo-IRR, demonstrating superior dereflection performance and robust generalization. The code is released at https://github.com/KONGBAI-8080/ReLo-IRR.
Chaoqun Wang, Yuehuan Wei, Haoxiang Cao +1
School of Artificial Intelligence, South China Normal University, Guangzhou, China · University of Science and Technology of China, Hefei, China