3D ViTs

ViT: Vision Transformer

Overview

3D Vision Transformers apply attention to volumetric images, point clouds, meshes, and multi-view images to learn spatial structure for tasks such as medical segmentation, object detection, and scene reconstruction. Common approaches include voxel or point-token encoders, query-based decoders, camera- or geometry-aware positional encodings, and fusion across views or modalities. They are evaluated on benchmarks including ScanNet, nuScenes, BraTS, and ShapeNet, with newer work also adapting pretrained geometry models to streaming video and long sequences. A central tension is how to preserve fine geometric detail and cross-view consistency while reducing the memory and computation required by attention, especially when transferring across sensors and domains.

Latest papers

157 papers13 papers in the last four weeks, down 19% on the four weeks before. 0.1% of all new papers.

Feb 25, 2026

Axial-Centric Cross-Plane Attention for 3D Medical Image Classification

Abridged: Clinicians commonly interpret 3D medical images by examining multiple anatomical planes rather than relying on volumetric views. In clinical CT workflows, the axial plane often serves as the primary diagnostic reference, while the auxiliary planes provide complementary spatial context. However, many existing 3D deep learning approaches either process volumetric data holistically or assign equal importance to all planes, failing to reflect this asymmetric, axial-centric interpretation strategy. To address this, we propose an axial-centric cross-plane attention architecture for 3D medical image classification that models asymmetric dependencies between anatomical planes. The architecture employs large-scale axial CT images pretrained MedDINOv3 as a frozen feature extractor for axial, coronal, and sagittal planes. RICA blocks and intra-plane transformer encoders capture plane-specific positional and contextual information, while axial-centric cross-plane transformer encoders selectively condition axial representations on complementary auxiliary representations. Experiments on six datasets from the MedMNIST3D benchmark show that the proposed method consistently outperforms existing 3D and multi-plane models in ACC and AUC. A lightweight variant, AC-Tiny, achieves competitive performance with substantially fewer trainable parameters, suggesting that architectural design contributes more to performance gains than increased model scale. Ablation studies further validate the importance of axial-centric querying, QKV allocation, directional cross-plane fusion, residual-free cross-attention, and classification head design. Slice-level Grad-CAM visualizations demonstrate that the model identifies diagnostically relevant regions across all planes. These findings highlight the value of aligning architectural design with clinical interpretation workflows for robust 3D medical image analysis.
Feb 24, 2026

Boosting Instance Awareness via Cross-View Correlation with 4D Radar and Camera for 3D Object Detection

4D millimeter-wave radar has emerged as a promising sensing modality for autonomous driving due to its robustness and affordability. However, its sparse and weak geometric cues make reliable instance activation difficult, limiting the effectiveness of existing radar-camera fusion paradigms. BEV-level fusion offers global scene understanding but suffers from weak instance focus, while perspective-level fusion captures instance details but lacks holistic context. To address these limitations, we propose SIFormer, a scene-instance aware transformer for 3D object detection using 4D radar and camera. SIFormer first suppresses background noise during view transformation through segmentation- and depth-guided localization. It then introduces a cross-view activation mechanism that injects 2D instance cues into BEV space, enabling reliable instance awareness under weak radar geometry. Finally, a transformer-based fusion module aggregates complementary image semantics and radar geometry for robust perception. As a result, with the aim of enhancing instance awareness, SIFormer bridges the gap between the two paradigms, combining their complementary strengths to address inherent sparse nature of radar and improve detection accuracy. Experiments demonstrate that SIFormer achieves state-of-the-art performance on View-of-Delft, TJ4DRadSet and NuScenes datasets. Source code is available at github.com/shawnnnkb/SIFormer.
Feb 18, 2026

RefineFormer3D: Efficient 3D Medical Image Segmentation via Adaptive Multi-Scale Transformer with Cross Attention Fusion

Accurate and computationally efficient 3D medical image segmentation remains a critical challenge in clinical workflows. Transformer-based architectures often demonstrate superior global contextual modeling but at the expense of excessive parameter counts and memory demands, restricting their clinical deployment. We propose RefineFormer3D, a lightweight hierarchical transformer architecture that balances segmentation accuracy and computational efficiency for volumetric medical imaging. The architecture integrates three key components: (i) GhostConv3D-based patch embedding for efficient feature extraction with minimal redundancy, (ii) MixFFN3D module with low-rank projections and depthwise convolutions for parameter-efficient feature extraction, and (iii) a cross-attention fusion decoder enabling adaptive multi-scale skip connection integration. RefineFormer3D contains only 2.94M parameters, substantially fewer than contemporary transformer-based methods. Extensive experiments on ACDC and BraTS benchmarks demonstrate that RefineFormer3D achieves 93.44% and 85.9% average Dice scores respectively, outperforming or matching state-of-the-art methods while requiring significantly fewer parameters. Furthermore, the model achieves fast inference (8.35 ms per volume on GPU) with low memory requirements, supporting deployment in resource-constrained clinical environments. These results establish RefineFormer3D as an effective and scalable solution for practical 3D medical image segmentation.
Oct 28, 2025

MIC-BEV: Multi-Infrastructure Camera Bird's-Eye-View Transformer with Relation-Aware Fusion for 3D Object Detection

Infrastructure-based perception plays a crucial role in intelligent transportation systems by providing global situational awareness and enabling cooperative autonomy. However, camera-based 3D object detection remains challenging under multi-view infrastructure setups, heterogeneous camera configurations, degraded sensing conditions, diverse road layouts, and adverse environmental conditions. We introduce MIC-BEV, a Transformer-based bird's-eye-view (BEV) perception framework for infrastructure-based multi-camera 3D object detection. MIC-BEV supports a variable number of cameras with heterogeneous intrinsic and extrinsic parameters and uses a graph-enhanced fusion module to adaptively integrate multi-view image features by modeling geometric relationships between cameras and the BEV grid. To support training and evaluation, we introduce M2I, a synthetic dataset featuring diverse camera layouts, road geometries, environmental conditions, and sensor-degradation cases. Experiments on the simulated M2I dataset and the real-world RoScenes dataset show that MIC-BEV achieves strong 3D detection performance and improved robustness under challenging conditions. Ablation studies validate the effectiveness of its key design choices, while weight analyses show that the learned fusion weights are associated with view-reliability factors and adapt to degraded camera inputs. We further report computational cost and deployment-related factors, providing additional evidence of practical feasibility and real-world deployment potential. These results indicate that MIC-BEV is a promising framework for infrastructure-based perception. The dataset and source code are available at: https://github.com/HandsomeYun/MIC-BEV.
Dec 17, 2024

CATSplat: Context-Aware Transformer with Spatial Guidance for Generalizable 3D Gaussian Splatting from A Single-View Image

Recently, generalizable feed-forward methods based on 3D Gaussian Splatting have gained significant attention for their potential to reconstruct 3D scenes using finite resources. These approaches create a 3D radiance field, parameterized by per-pixel 3D Gaussian primitives, from just a few images in a single forward pass. However, unlike multi-view methods that benefit from cross-view correspondences, 3D scene reconstruction with a single-view image remains an underexplored area. In this work, we introduce CATSplat, a novel generalizable transformer-based framework designed to break through the inherent constraints in monocular settings. First, we propose leveraging textual guidance from a visual-language model to complement insufficient information from a single image. By incorporating scene-specific contextual details from text embeddings through cross-attention, we pave the way for context-aware 3D scene reconstruction beyond relying solely on visual cues. Moreover, we advocate utilizing spatial guidance from 3D point features toward comprehensive geometric understanding under single-view settings. With 3D priors, image features can capture rich structural insights for predicting 3D Gaussians without multi-view techniques. Extensive experiments on large-scale datasets demonstrate the state-of-the-art performance of CATSplat in single-view 3D scene reconstruction with high-quality novel view synthesis.
Sep 20, 2024

Tackling fluffy clouds: robust agricultural field boundary delineation from Sentinel-1 and Sentinel-2 satellite image time series

Accurate delineation of agricultural field boundaries is essential for effective crop monitoring and resource management. However, competing methodologies often face significant challenges, particularly in their reliance on extensive manual efforts for cloud-free data curation and limited adaptability to diverse global conditions. In this paper, we introduce PTAViT3D, a deep learning architecture specifically designed for processing three-dimensional time series of satellite imagery from either Sentinel-1 (S1) or Sentinel-2 (S2). Additionally, we present PTAViT3D-CA, an extension of the PTAViT3D model incorporating cross-attention mechanisms to fuse S1 and S2 datasets, enhancing robustness in cloud-contaminated scenarios. The proposed methods leverage spatio-temporal correlations through a memory-efficient 3D Vision Transformer architecture, facilitating accurate boundary delineation directly from preprocessed, cloud-affected imagery. We comprehensively validate our models through extensive testing on various datasets, including Australia's ePaddocks - CSIRO's national, continental-scale agricultural field boundary product covering Australia's cropping regions - alongside public benchmarks Fields-of-the-World, PASTIS, and AI4SmallFarms. Our results consistently demonstrate state-of-the-art performance, highlighting excellent global transferability and robustness. Crucially, our approach significantly simplifies data preparation workflows by reliably processing cloud-affected imagery, thereby offering strong adaptability across diverse agricultural environments. Our code and models are publicly available at https://github.com/feevos/tfcl.
Aug 20, 2024

Multi-view Hand Reconstruction with a Point-Embedded Transformer

This work introduces a novel and generalizable multi-view Hand Mesh Reconstruction (HMR) model, named POEM, designed for practical use in real-world hand motion capture scenarios. The advances of the POEM model consist of two main aspects. First, concerning the modeling of the problem, we propose embedding a static basis point within the multi-view stereo space. A point represents a natural form of 3D information and serves as an ideal medium for fusing features across different views, given its varied projections across these views. Consequently, our method harnesses a simple yet effective idea: a complex 3D hand mesh can be represented by a set of 3D basis points that 1) are embedded in the multi-view stereo, 2) carry features from the multi-view images, and 3) encompass the hand in it. The second advance lies in the training strategy. We utilize a combination of five large-scale multi-view datasets and employ randomization in the number, order, and poses of the cameras. By processing such a vast amount of data and a diverse array of camera configurations, our model demonstrates notable generalizability in the real-world applications. As a result, POEM presents a highly practical, plug-and-play solution that enables user-friendly, cost-effective multi-view motion capture for both left and right hands. The model and source codes are available at https://github.com/JubSteven/POEM-v2.