UniqueSplat: View-conditioned 3D Gaussian Splatting for Generalizable 3D Reconstruction
2026-08-03 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial Intelligence
AI summaryⓘ
The authors introduce UniqueSplat, a method that creates 3D images that change depending on the viewpoint you're looking from. Unlike previous methods that use a fixed set of visual elements for all views, UniqueSplat adjusts these elements for each specific view by learning information about the view itself. They use a special network with two parts: one that learns common features across all views and another that focuses on individual views. Their tests show that UniqueSplat performs better than existing models and works well even on new datasets it wasn't trained on.
3D Gaussian SplattingRadiance FieldsView-Conditioned ModelsFeed-Forward NetworksHyperNetworkView-Agnostic EmbeddingsCross-Dataset EvaluationNeural RenderingRealEstate10KDTU Dataset
Authors
Haixu Song, Xiaoke Yang, Shengjun Zhang, Jiwen Lu, Yueqi Duan
Abstract
In this paper, we propose UniqueSplat, a view-conditioned feed-forward 3D Gaussian Splatting model to reconstruct customized 3D radiance fields for each view query. Existing feed-forward methods such as pixelSplat and MVSplat aim to generate fixed Gaussians across all views of each scene by minimizing the error between rendered views and ground-truth images. However, such fixed Gaussians generally render images from all views and lack the ability to adapt to specific viewpoints, as they do not incorporate target view information when predicting Gaussians. To address this, our UniqueSplat learns the view-conditioned information as a prior and incorporates this knowledge into network parameters, so that Gaussians are dynamically adjusted in accordance with different views. Specifically, we propose a two-branch view-conditioned hyperNetwork to simultaneously learn view-agnostic embeddings and view-specific knowledge, which not only explores the shareable knowledge from various views, but also adapts the model to specific views at test time. Extensive experiments on widely-used datasets including RealEstate10K, ACID and DTU demonstrate the superiority of UniqueSplat over the state-of-the-art methods. Moreover, UniqueSplat encouragingly outperforms existing methods in cross-dataset evaluation, showing its notable generalization ability.