Dehaze-then-Splat uses per-frame generative dehazing followed by physics-regularized 3D Gaussian Splatting to achieve 20.98 dB PSNR and 0.683 SSIM on the Akikaze scene, a 1.5 dB gain over baseline by mitigating cross-view inconsistencies.
Training-Free Model Ensemble for Single-Image Super-Resolution via Strong-Branch Compensation
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
abstract
Single-image super-resolution has progressed from deep convolutional baselines to stronger Transformer and state-space architectures, yet the corresponding performance gains typically come with higher training cost, longer engineering iteration, and heavier deployment burden. In many practical settings, multiple pretrained models with partially complementary behaviors are already available, and the binding constraint is no longer architectural capacity but how effectively their outputs can be combined without additional training. Rather than pursuing further architectural redesign, this paper proposes a training-free output-level ensemble framework. A dual-branch pipeline is constructed in which a Hybrid attention network with TLC inference provides stable main reconstruction, while a MambaIRv2 branch with geometric self-ensemble supplies strong compensation for high-frequency detail recovery. The two branches process the same low-resolution input independently and are fused in the image space via a lightweight weighted combination, without updating any model parameters or introducing an additional trainable module. As our solution to the NTIRE 2026 Image Super-Resolution ($\times 4$) Challenge, the proposed design consistently improves over the base branch and slightly exceeds the pure strong branch in PSNR at the best operating point under a unified DIV2K bicubic $\times 4$ evaluation protocol. Ablation studies confirm that output-level compensation provides a low-overhead and practically accessible upgrade path for existing super-resolution systems.
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 7years
2026 7roles
background 1polarities
unclear 1representative citing papers
A framework that combines MLLM-based image enhancement with a medium-aware 3D Gaussian Splatting model to reconstruct and render smoke scenes.
CLIP-guided selection of external data plus staged NAFNet training and inference fusion provides an effective pipeline for nighttime image dehazing in the NTIRE 2026 challenge.
Dual-branch fusion of HAT-L and MambaIRv2-L with eight-way ensemble and equal-weight averaging outperforms single branches on PSNR, SSIM, and challenge score for infrared super-resolution.
SmokeGS-R uses refined dark channel prior for pseudo-clean supervision to train 3DGS geometry, followed by ensemble-based appearance harmonization, achieving PSNR 15.21 and outperforming baselines on smoke restoration challenge data.
Expanding training data diversity, adopting two-stage optimization, and applying geometric self-ensemble raises Restormer performance on Gaussian color denoising at sigma=50 by 3.366 dB PSNR on the NTIRE 2026 validation set.
The NTIRE 2026 challenge reports measurable progress in 3D reconstruction pipelines that handle real-world low-light and smoke degradation via the RealX3D benchmark.
citing papers explorer
-
Dehaze-then-Splat: Generative Dehazing with Physics-Informed 3D Gaussian Splatting for Smoke-Free Novel View Synthesis
Dehaze-then-Splat uses per-frame generative dehazing followed by physics-regularized 3D Gaussian Splatting to achieve 20.98 dB PSNR and 0.683 SSIM on the Akikaze scene, a 1.5 dB gain over baseline by mitigating cross-view inconsistencies.
-
3D Smoke Scene Reconstruction Guided by Vision Priors from Multimodal Large Language Models
A framework that combines MLLM-based image enhancement with a medium-aware 3D Gaussian Splatting model to reconstruct and render smoke scenes.
-
CLIP-Guided Data Augmentation for Night-Time Image Dehazing
CLIP-guided selection of external data plus staged NAFNet training and inference fusion provides an effective pipeline for nighttime image dehazing in the NTIRE 2026 challenge.
-
Dual-Branch Remote Sensing Infrared Image Super-Resolution
Dual-branch fusion of HAT-L and MambaIRv2-L with eight-way ensemble and equal-weight averaging outperforms single branches on PSNR, SSIM, and challenge score for infrared super-resolution.
-
SmokeGS-R: Physics-Guided Pseudo-Clean 3DGS for Real-World Multi-View Smoke Restoration
SmokeGS-R uses refined dark channel prior for pseudo-clean supervision to train 3DGS geometry, followed by ensemble-based appearance harmonization, achieving PSNR 15.21 and outperforming baselines on smoke restoration challenge data.
-
Beyond Model Design: Data-Centric Training and Self-Ensemble for Gaussian Color Image Denoising
Expanding training data diversity, adopting two-stage optimization, and applying geometric self-ensemble raises Restormer performance on Gaussian color denoising at sigma=50 by 3.366 dB PSNR on the NTIRE 2026 validation set.
-
NTIRE 2026 3D Restoration and Reconstruction in Real-world Adverse Conditions: RealX3D Challenge Results
The NTIRE 2026 challenge reports measurable progress in 3D reconstruction pipelines that handle real-world low-light and smoke degradation via the RealX3D benchmark.