Context and Geometry Aware Voxel Transformer for Semantic Scene Completion

• URL: http://arxiv.org/abs/2405.13675v4
• Date: Tue, 22 Oct 2024 14:28:56 GMT
• Title: Context and Geometry Aware Voxel Transformer for Semantic Scene Completion
• Authors: Zhu Yu, Runmin Zhang, Jiacheng Ying, Junchen Yu, Xiaohai Hu, Lun Luo, Si-Yuan Cao, Hui-Liang Shen,
• Abstract summary: Vision-based Semantic Scene Completion (SSC) has gained much attention due to its widespread applications in various 3D perception tasks. Existing sparse-to-dense approaches typically employ shared context-independent queries across various input images. We introduce a neural network named CGFormer to achieve semantic scene completion.
• Score: 7.147020285382786
• License:
• Abstract: Vision-based Semantic Scene Completion (SSC) has gained much attention due to its widespread applications in various 3D perception tasks. Existing sparse-to-dense approaches typically employ shared context-independent queries across various input images, which fails to capture distinctions among them as the focal regions of different inputs vary and may result in undirected feature aggregation of cross-attention. Additionally, the absence of depth information may lead to points projected onto the image plane sharing the same 2D position or similar sampling points in the feature map, resulting in depth ambiguity. In this paper, we present a novel context and geometry aware voxel transformer. It utilizes a context aware query generator to initialize context-dependent queries tailored to individual input images, effectively capturing their unique characteristics and aggregating information within the region of interest. Furthermore, it extend deformable cross-attention from 2D to 3D pixel space, enabling the differentiation of points with similar image coordinates based on their depth coordinates. Building upon this module, we introduce a neural network named CGFormer to achieve semantic scene completion. Simultaneously, CGFormer leverages multiple 3D representations (i.e., voxel and TPV) to boost the semantic and geometric representation abilities of the transformed 3D volume from both local and global perspectives. Experimental results demonstrate that CGFormer achieves state-of-the-art performance on the SemanticKITTI and SSCBench-KITTI-360 benchmarks, attaining a mIoU of 16.87 and 20.05, as well as an IoU of 45.99 and 48.07, respectively. Remarkably, CGFormer even outperforms approaches employing temporal images as inputs or much larger image backbone networks.

Related papers

• Large Spatial Model: End-to-end Unposed Images to Semantic 3D [79.94479633598102]
  Large Spatial Model (LSM) processes unposed RGB images directly into semantic radiance fields. LSM simultaneously estimates geometry, appearance, and semantics in a single feed-forward operation. It can generate versatile label maps by interacting with language at novel viewpoints.
  arXiv  Detail & Related papers  (2024-10-24T17:54:42Z)
• Multiview Scene Graph [7.460438046915524]
  A proper scene representation is central to the pursuit of spatial intelligence. We propose to build Multiview Scene Graphs (MSG) from unposed images. MSG represents a scene topologically with interconnected place and object nodes.
  arXiv  Detail & Related papers  (2024-10-15T02:04:05Z)
• Monocular Visual Place Recognition in LiDAR Maps via Cross-Modal State Space Model and Multi-View Matching [2.400446821380503]
  We introduce an efficient framework to learn descriptors for both RGB images and point clouds. It takes visual state space model (VMamba) as the backbone and employs a pixel-view-scene joint training strategy. A visible 3D points overlap strategy is then designed to quantify the similarity between point cloud views and RGB images for multi-view supervision.
  arXiv  Detail & Related papers  (2024-10-08T18:31:41Z)
• Boosting Cross-Domain Point Classification via Distilling Relational Priors from 2D Transformers [59.0181939916084]
  Traditional 3D networks mainly focus on local geometric details and ignore the topological structure between local geometries. We propose a novel Priors Distillation (RPD) method to extract priors from the well-trained transformers on massive images. Experiments on the PointDA-10 and the Sim-to-Real datasets verify that the proposed method consistently achieves the state-of-the-art performance of UDA for point cloud classification.
  arXiv  Detail & Related papers  (2024-07-26T06:29:09Z)
• SeMLaPS: Real-time Semantic Mapping with Latent Prior Networks and Quasi-Planar Segmentation [53.83313235792596]
  We present a new methodology for real-time semantic mapping from RGB-D sequences. It combines a 2D neural network and a 3D network based on a SLAM system with 3D occupancy mapping. Our system achieves state-of-the-art semantic mapping quality within 2D-3D networks-based systems.
  arXiv  Detail & Related papers  (2023-06-28T22:36:44Z)
• FrustumFormer: Adaptive Instance-aware Resampling for Multi-view 3D Detection [47.6570523164125]
  We propose a novel framework named FrustumFormer, which pays more attention to the features in instance regions via adaptive instance-aware resampling. Experiments on the nuScenes dataset demonstrate the effectiveness of FrustumFormer, and we achieve a new state-of-the-art performance on the benchmark.
  arXiv  Detail & Related papers  (2023-01-10T17:51:55Z)
• Multi-Projection Fusion and Refinement Network for Salient Object Detection in 360{\deg} Omnidirectional Image [141.10227079090419]
  We propose a Multi-Projection Fusion and Refinement Network (MPFR-Net) to detect the salient objects in 360deg omnidirectional image. MPFR-Net uses the equirectangular projection image and four corresponding cube-unfolding images as inputs. Experimental results on two omnidirectional datasets demonstrate that the proposed approach outperforms the state-of-the-art methods both qualitatively and quantitatively.
  arXiv  Detail & Related papers  (2022-12-23T14:50:40Z)
• Vision Transformer for NeRF-Based View Synthesis from a Single Input Image [49.956005709863355]
  We propose to leverage both the global and local features to form an expressive 3D representation. To synthesize a novel view, we train a multilayer perceptron (MLP) network conditioned on the learned 3D representation to perform volume rendering. Our method can render novel views from only a single input image and generalize across multiple object categories using a single model.
  arXiv  Detail & Related papers  (2022-07-12T17:52:04Z)
• Similarity-Aware Fusion Network for 3D Semantic Segmentation [87.51314162700315]
  We propose a similarity-aware fusion network (SAFNet) to adaptively fuse 2D images and 3D point clouds for 3D semantic segmentation. We employ a late fusion strategy where we first learn the geometric and contextual similarities between the input and back-projected (from 2D pixels) point clouds. We show that SAFNet significantly outperforms existing state-of-the-art fusion-based approaches across various data integrity.
  arXiv  Detail & Related papers  (2021-07-04T09:28:18Z)

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.