Search papers, labs, and topics across Lattice.
Dehaze-then-Splat is a two-stage pipeline for smoke removal and novel view synthesis, first applying per-frame generative dehazing with Nano Banana Pro, followed by training 3D Gaussian Splatting (3DGS). To address inconsistencies from per-frame dehazing, the method incorporates physics-informed auxiliary losses, including depth supervision and dark channel prior regularization. The approach mitigates artifacts via MCMC-based densification with early stopping and haze-suppression priors, achieving a +1.50dB PSNR improvement over the baseline on the Akikaze validation scene.
Frame-by-frame image enhancement can actually *hurt* 3D reconstruction, but Dehaze-then-Splat shows how physics-informed regularization can turn this liability into an asset for novel view synthesis.
We present Dehaze-then-Splat, a two-stage pipeline for multi-view smoke removal and novel view synthesis developed for Track~2 of the NTIRE 2026 3D Restoration and Reconstruction Challenge. In the first stage, we produce pseudo-clean training images via per-frame generative dehazing using Nano Banana Pro, followed by brightness normalization. In the second stage, we train 3D Gaussian Splatting (3DGS) with physics-informed auxiliary losses -- depth supervision via Pearson correlation with pseudo-depth, dark channel prior regularization, and dual-source gradient matching -- that compensate for cross-view inconsistencies inherent in frame-wise generative processing. We identify a fundamental tension in dehaze-then-reconstruct pipelines: per-image restoration quality does not guarantee multi-view consistency, and such inconsistency manifests as blurred renders and structural instability in downstream 3D reconstruction.Our analysis shows that MCMC-based densification with early stopping, combined with depth and haze-suppression priors, effectively mitigates these artifacts. On the Akikaze validation scene, our pipeline achieves 20.98\,dB PSNR and 0.683 SSIM for novel view synthesis, a +1.50\,dB improvement over the unregularized baseline.
