novel-view-synthesis

HRoPE introduces a surface-aware positional encoding that models each patch as a slanted 3D plane and injects plane-induced homographies as relative transformations into attention, improving cross-view alignment and novel view synthesis without ground-truth geometry. Experiments on RealEstate10K, CO3Dv2, and DL3DV show it outperforms existing geometric positional encodings.

FineSplat 提出一种基于细粒度图像 patch 的前馈 3D 高斯溅射框架，通过双分组策略、渲染反馈和紧凑高斯生成，在更小的高斯预算下实现了更精细的几何与外观重建并达到 SOTA 结果。

The paper introduces Seed Injection and Temporal Biased Relocation to allocate Gaussian primitives based on predicted spatiotemporal destinations rather than current locations, improving reconstruction of small fast-moving objects and fine textures in dynamic scenes without increasing the Gaussian budget.

InfiniSplat v2 通过解耦几何与属性预测、使用真实相机位姿对齐以及大角度视角监督，显著提升前馈式 3D 高斯重建的新视角渲染质量，在六个数据集上取得 SOTA 的 LPIPS 和 DISTS 指标。

4LC 提出一种用于动态世界建模的 4D 潜空间缓存，通过复合视觉表示（VAE、DINO、V-JEPA、VGGT）和灵活的时空访问接口，高效复用静态与动态内容作为生成条件，支持相机控制、时间控制、新视角合成与长时生成等多项视频生成任务，且无需任务特定微调。

CompactSplat is a feedforward framework that adaptively allocates 3D Gaussians via texture-guided spatial partitioning and hierarchical sampling, improving geometry and texture recovery in sparse-view scene reconstruction while supporting dynamic sparsity control without retraining.

LEGO 提出一种无提升（lifting-free）的方法，用 LVSM 风格的 Transformer 直接合成第一人称视角，替代传统的深度估计与点云重投影管线，并将结果作为条件输入视频扩散模型，在跨视角视频生成上超越现有 SOTA 且能跨数据集泛化。

OuroWorld 是一个无掩码框架，可将任意静态 3D Gaussian Splatting 场景转化为从任意视角无缝循环的 3D 动态摄影（cinemagraph），利用视觉语言模型引导视频模型生成参考视频，并提出 Inconsistency-Robust Periodic 4DGS 与 Grounded Drift Field 来保证循环一致性并吸收跨视角不一致，在用户研究中以 70.8%-99.0% 的胜率超越所有基线。

Mobile-4DGS 提出了一个统一的轻量级框架，在资源受限的移动设备上实现高保真、实时的静态与动态高斯溅射渲染。该方法通过压缩外观建模、多视角致密化与剪枝，以及紧凑的显式4D表示（含静态-动态划分与深度顺序复用），显著降低了存储和渲染开销，同时保持了有竞争力的视觉质量。

NAMVIS proposes a diffusion-free framework that reformulates sparse-view multi-view image synthesis as geometry-conditioned next-scale autoregression, using Multi-scale Projective Pose Encoding to anchor generation to camera geometry. It outperforms diffusion baselines on Objaverse, GSO, and OmniObject3D in PSNR/SSIM/LPIPS while running over 3x faster.

DyRAD presents a radar novel view synthesis method for dynamic driving scenes that models static background and motion-tracked dynamic point reflectors to render complete range-azimuth-Doppler tensors, using a fixed analytic point-spread function to enable zero-shot sensor-configuration transfer and outperforming baselines by a wide margin on the RADIal benchmark.

GeoVerse is a framework for synthesizing world-consistent novel views by performing generation in a geometric latent space from a 3D foundation model and injecting appearance priors from a video generative model, demonstrating improved visual quality and geometric consistency over existing methods.

VGGT-Diff is a geometry-routed multi-view diffusion model that integrates visual geometry latents with a pretrained diffusion model to improve sparse-view novel view synthesis, achieving competitive or state-of-the-art performance.

ATSplat introduces a feed-forward 3D Gaussian Splatting framework that uses adaptive 3D tokens to allocate primitives based on scene complexity, achieving state-of-the-art rendering quality while reducing the number of Gaussians by over 5.7 times compared to dense methods.

AsySplat proposes an asymmetric architecture that decouples geometry and appearance modeling in 3D Gaussian Splatting, achieving high efficiency for long-sequence scene modeling with nearly 800x speedup over optimization-based methods.

MetaView proposes a diffusion-based monocular novel view synthesis framework that combines implicit geometry priors with metric depth guidance to achieve consistent and controllable rendering under large viewpoint changes from a single image.

Flux-GS enables real-time high-fidelity 3D Gaussian Splatting on mobile platforms through efficient lighting representation, attribute-conditioned enhancement, and multi-view densification strategies.

Introduces DF3DV-1K, a large-scale real-world dataset with 1,048 scenes and 89,924 images for distractor-free novel view synthesis, along with a benchmark of nine methods and an application improving radiance field methods via fine-tuning a diffusion-based 2D enhancer.

Track2View generates novel camera viewpoints from videos by conditioning a video diffusion transformer on paired 3D point tracks, achieving state-of-the-art visual quality and significant reductions in rotation and translation errors.

A training-free framework for spatial reasoning from egocentric videos that enables revisiting conclusions through synthesized novel-view videos generated from predicted 3D geometry.