EmbodiedSplat: Online Feed-Forward Semantic 3DGS for Open-Vocabulary 3D Scene Understanding

• URL: http://arxiv.org/abs/2603.04254v1
• Date: Wed, 04 Mar 2026 16:40:41 GMT
• Title: EmbodiedSplat: Online Feed-Forward Semantic 3DGS for Open-Vocabulary 3D Scene Understanding
• Authors: Seungjun Lee, Zihan Wang, Yunsong Wang, Gim Hee Lee,
• Abstract summary: EmbodiedSplat is an online feed-forward 3DGS for open-vocabulary scene understanding.<n>Our objectives are 1) Reconstructs the semantic-embedded 3DGS of the entire scene from over 300 streaming images in an online manner, and 2) Highly generalizable to novel scenes with feed-forward design.
• Score: 66.80528512321106
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Understanding a 3D scene immediately with its exploration is essential for embodied tasks, where an agent must construct and comprehend the 3D scene in an online and nearly real-time manner. In this study, we propose EmbodiedSplat, an online feed-forward 3DGS for open-vocabulary scene understanding that enables simultaneous online 3D reconstruction and 3D semantic understanding from the streaming images. Unlike existing open-vocabulary 3DGS methods which are typically restricted to either offline or per-scene optimization setting, our objectives are two-fold: 1) Reconstructs the semantic-embedded 3DGS of the entire scene from over 300 streaming images in an online manner. 2) Highly generalizable to novel scenes with feed-forward design and supports nearly real-time 3D semantic reconstruction when combined with real-time 2D models. To achieve these objectives, we propose an Online Sparse Coefficients Field with a CLIP Global Codebook where it binds the 2D CLIP embeddings to each 3D Gaussian while minimizing memory consumption and preserving the full semantic generalizability of CLIP. Furthermore, we generate 3D geometric-aware CLIP features by aggregating the partial point cloud of 3DGS through 3D U-Net to compensate the 3D geometric prior to 2D-oriented language embeddings. Extensive experiments on diverse indoor datasets, including ScanNet, ScanNet++, and Replica, demonstrate both the effectiveness and efficiency of our method. Check out our project page in https://0nandon.github.io/EmbodiedSplat/.

Related papers

• Unified Semantic Transformer for 3D Scene Understanding [55.415468022487005]
  We introduce UNITE, a novel feed-forward neural network that unifies a diverse set of 3D semantic tasks within a single model.<n>Our model operates on unseen scenes in a fully end-to-end manner and only takes a few seconds to infer the full 3D semantic geometry.<n>We demonstrate that UNITE achieves state-of-the-art performance on several different semantic tasks and even outperforms task-specific models.
  arXiv  Detail & Related papers  (2025-12-16T12:49:35Z)
• EA3D: Online Open-World 3D Object Extraction from Streaming Videos [55.48835711373918]
  We present ExtractAnything3D (EA3D), a unified online framework for open-world 3D object extraction.<n>Given a streaming video, EA3D dynamically interprets each frame using vision-language and 2D vision foundation encoders to extract object-level knowledge.<n>A recurrent joint optimization module directs the model's attention to regions of interest, simultaneously enhancing both geometric reconstruction and semantic understanding.
  arXiv  Detail & Related papers  (2025-10-29T03:56:41Z)
• Dr. Splat: Directly Referring 3D Gaussian Splatting via Direct Language Embedding Registration [41.046653227409564]
  Dr. Splat is a novel approach for open-vocabulary 3D scene understanding leveraging 3D Gaussian Splatting.<n>Our method associates language-aligned CLIP embeddings with 3D Gaussians for holistic 3D scene understanding.<n> Experiments demonstrate that our approach significantly outperforms existing approaches in 3D perception benchmarks.
  arXiv  Detail & Related papers  (2025-02-23T17:01:14Z)
• PanoSLAM: Panoptic 3D Scene Reconstruction via Gaussian SLAM [105.01907579424362]
  PanoSLAM is the first SLAM system to integrate geometric reconstruction, 3D semantic segmentation, and 3D instance segmentation within a unified framework.<n>For the first time, it achieves panoptic 3D reconstruction of open-world environments directly from the RGB-D video.
  arXiv  Detail & Related papers  (2024-12-31T08:58:10Z)
• Rethinking Open-Vocabulary Segmentation of Radiance Fields in 3D Space [10.49905491984899]
  We redefine the problem to segment the 3D volume and propose the following methods for better 3D understanding.<n>We directly supervise the 3D points to train the language embedding field, unlike previous methods that anchor supervision at 2D pixels.<n>We transfer the learned language field to 3DGS, achieving the first real-time rendering speed without sacrificing training time or accuracy.
  arXiv  Detail & Related papers  (2024-08-14T09:50:02Z)
• ALSTER: A Local Spatio-Temporal Expert for Online 3D Semantic Reconstruction [62.599588577671796]
  We propose an online 3D semantic segmentation method that incrementally reconstructs a 3D semantic map from a stream of RGB-D frames. Unlike offline methods, ours is directly applicable to scenarios with real-time constraints, such as robotics or mixed reality.
  arXiv  Detail & Related papers  (2023-11-29T20:30:18Z)
• SSR-2D: Semantic 3D Scene Reconstruction from 2D Images [54.46126685716471]
  In this work, we explore a central 3D scene modeling task, namely, semantic scene reconstruction without using any 3D annotations. The key idea of our approach is to design a trainable model that employs both incomplete 3D reconstructions and their corresponding source RGB-D images. Our method achieves the state-of-the-art performance of semantic scene completion on two large-scale benchmark datasets MatterPort3D and ScanNet.
  arXiv  Detail & Related papers  (2023-02-07T17:47:52Z)
• Learning 3D Scene Priors with 2D Supervision [37.79852635415233]
  We propose a new method to learn 3D scene priors of layout and shape without requiring any 3D ground truth. Our method represents a 3D scene as a latent vector, from which we can progressively decode to a sequence of objects characterized by their class categories. Experiments on 3D-FRONT and ScanNet show that our method outperforms state of the art in single-view reconstruction.
  arXiv  Detail & Related papers  (2022-11-25T15:03:32Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.