TokenSplat: Token-aligned 3D Gaussian Splatting for Feed-forward Pose-free Reconstruction

TL;DR

TokenSplat is a novel feed-forward framework that jointly reconstructs 3D scenes and estimates camera poses from unposed multi-view images, using token-aligned Gaussian prediction and specialized decoders for improved accuracy.

Contribution

It introduces a new token-aligned Gaussian prediction module and an asymmetric dual-flow decoder, enabling pose-free 3D reconstruction and pose estimation within a purely feed-forward architecture.

Findings

Achieves higher reconstruction fidelity and novel-view synthesis quality.

Significantly improves pose estimation accuracy over prior pose-free methods.

Operates without iterative refinement, ensuring efficiency.

Abstract

We present TokenSplat, a feed-forward framework for joint 3D Gaussian reconstruction and camera pose estimation from unposed multi-view images. At its core, TokenSplat introduces a Token-aligned Gaussian Prediction module that aligns semantically corresponding information across views directly in the feature space. Guided by coarse token positions and fusion confidence, it aggregates multi-scale contextual features to enable long-range cross-view reasoning and reduce redundancy from overlapping Gaussians. To further enhance pose robustness and disentangle viewpoint cues from scene semantics, TokenSplat employs learnable camera tokens and an Asymmetric Dual-Flow Decoder (ADF-Decoder) that enforces directionally constrained communication between camera and image tokens. This maintains clean factorization within a feed-forward architecture, enabling coherent reconstruction and stable pose estimation without iterative refinement. Extensive experiments demonstrate that TokenSplat achieves higher reconstruction fidelity and novel-view synthesis quality in pose-free settings, and significantly improves pose estimation accuracy compared to prior pose-free methods. Project page: https://kidleyh.github.io/tokensplat/.