Weixun Luo

Weixun Luo, Alexandre Triay Bagur, Paul Aljabar et al.

Background and Objective: The lack of benchmark datasets has impeded the development of slice-to-volume registration algorithms. Such datasets are difficult to annotate, primarily due to the dimensional difference within data and the dearth of task-specific software. We aim to develop a user-friendly tool to streamline dataset annotation for slice-to-volume registration. Methods: The proposed tool, named SVRDA, is an installation-free web application for platform-agnostic collaborative dataset annotation. It enables efficient transformation manipulation via keyboard shortcuts and smooth case transitions with auto-saving. SVRDA supports configuration-based data loading and adheres to the separation of concerns, offering great flexibility and extensibility for future research. Various supplementary features have been implemented to facilitate slice-to-volume registration. Results: We validated the effectiveness of SVRDA by indirectly evaluating the post-registration segmentation quality on UK Biobank data, observing a dramatic overall improvement (24.02% in the Dice Similarity Coefficient and 48.93% in the 95th percentile Hausdorff distance, respectively) supported by highly statistically significant evidence ($p<0.001$).We further showcased the clinical usage of SVRDA by integrating it into test-retest T1 quantification on in-house magnetic resonance images, leading to more consistent results after registration. Conclusions: SVRDA can facilitate collaborative annotation of benchmark datasets while being potentially applicable to other pipelines incorporating slice-to-volume registration. Full source code and documentation are available at https://github.com/Roldbach/SVRDA

Weixun Luo, Xiaodan Xing, Guang Yang

Featured by a bottleneck structure, autoencoder (AE) and its variants have been largely applied in various medical image analysis tasks, such as segmentation, reconstruction and de-noising. Despite of their promising performances in aforementioned tasks, in this paper, we claim that AE models are not applicable to single image super-resolution (SISR) for 3D CT data. Our hypothesis is that the bottleneck architecture that resizes feature maps in AE models degrades the details of input images, thus can sabotage the performance of super-resolution. Although U-Net proposed skip connections that merge information from different levels, we claim that the degrading impact of feature resizing operations could hardly be removed by skip connections. By conducting large-scale ablation experiments and comparing the performance between models with and without the bottleneck design on a public CT lung dataset , we have discovered that AE models, including U-Net, have failed to achieve a compatible SISR result ($p<0.05$ by Student's t-test) compared to the baseline model. Our work is the first comparative study investigating the suitability of AE architecture for 3D CT SISR tasks and brings a rationale for researchers to re-think the choice of model architectures especially for 3D CT SISR tasks. The full implementation and trained models can be found at: https://github.com/Roldbach/Autoencoder-3D-CT-SISR

Ye Mao, Weixun Luo, Ranran Huang et al.

Pretraining 3D encoders by aligning with Contrastive Language Image Pretraining (CLIP) has emerged as a promising direction to learn generalizable representations for 3D scene understanding. In this paper, we propose UniScene3D, a transformer-based encoder that learns unified scene representations from multi-view colored pointmaps, jointly modeling image appearance and geometry. For robust colored pointmap representation learning, we introduce novel cross-view geometric alignment and grounded view alignment to enforce cross-view geometry and semantic consistency. Extensive low-shot and task-specific fine-tuning evaluations on viewpoint grounding, scene retrieval, scene type classification, and 3D VQA demonstrate our state-of-the-art performance. These results highlight the effectiveness of our approach for unified 3D scene understanding. https://yebulabula.github.io/UniScene3D/

Ranran Huang, Weixun Luo, Ye Mao et al.

In this paper, we introduce NAS3R, a self-supervised feed-forward framework that jointly learns explicit 3D geometry and camera parameters with no ground-truth annotations and no pretrained priors. During training, NAS3R reconstructs 3D Gaussians from uncalibrated and unposed context views and renders target views using its self-predicted camera parameters, enabling self-supervised training from 2D photometric supervision. To ensure stable convergence, NAS3R integrates reconstruction and camera prediction within a shared transformer backbone regulated by masked attention, and adopts a depth-based Gaussian formulation that facilitates well-conditioned optimization. The framework is compatible with state-of-the-art supervised 3D reconstruction architectures and can incorporate pretrained priors or intrinsic information when available. Extensive experiments show that NAS3R achieves superior results to other self-supervised methods, establishing a scalable and geometry-aware paradigm for 3D reconstruction from unconstrained data. Code and models are publicly available at https://ranrhuang.github.io/nas3r/.

Ye Mao, Weixun Luo, Junpeng Jing et al.

The rise of vision-language foundation models marks an advancement in bridging the gap between human and machine capabilities in 3D scene reasoning. Existing 3D reasoning benchmarks assume real-time scene accessibility, which is impractical due to the high cost of frequent scene updates. To this end, we introduce Hypothetical 3D Reasoning, namely Hypo3D, a benchmark designed to evaluate models' ability to reason without access to real-time scene data. Models need to imagine the scene state based on a provided change description before reasoning. Hypo3D is formulated as a 3D Visual Question Answering (VQA) benchmark, comprising 7,727 context changes across 700 indoor scenes, resulting in 14,885 question-answer pairs. An anchor-based world frame is established for all scenes, ensuring consistent reference to a global frame for directional terms in context changes and QAs. Extensive experiments show that state-of-the-art foundation models struggle to reason in hypothetically changed scenes. This reveals a substantial performance gap compared to humans, particularly in scenarios involving movement changes and directional reasoning. Even when the context change is irrelevant to the question, models often incorrectly adjust their answers. Project website: https://matchlab-imperial.github.io/Hypo3D/

Junpeng Jing, Weixun Luo, Ye Mao et al.

This paper introduces Stereo Any Video, a powerful framework for video stereo matching. It can estimate spatially accurate and temporally consistent disparities without relying on auxiliary information such as camera poses or optical flow. The strong capability is driven by rich priors from monocular video depth models, which are integrated with convolutional features to produce stable representations. To further enhance performance, key architectural innovations are introduced: all-to-all-pairs correlation, which constructs smooth and robust matching cost volumes, and temporal convex upsampling, which improves temporal coherence. These components collectively ensure robustness, accuracy, and temporal consistency, setting a new standard in video stereo matching. Extensive experiments demonstrate that our method achieves state-of-the-art performance across multiple datasets both qualitatively and quantitatively in zero-shot settings, as well as strong generalization to real-world indoor and outdoor scenarios.

Ye Mao, Weixun Luo, Ranran Huang et al.

In this paper, we introduce POMA-3D, the first self-supervised 3D representation model learned from point maps. Point maps encode explicit 3D coordinates on a structured 2D grid, preserving global 3D geometry while remaining compatible with the input format of 2D foundation models. To transfer rich 2D priors into POMA-3D, a view-to-scene alignment strategy is designed. Moreover, as point maps are view-dependent with respect to a canonical space, we introduce POMA-JEPA, a joint embedding-predictive architecture that enforces geometrically consistent point map features across multiple views. Additionally, we introduce ScenePoint, a point map dataset constructed from 6.5K room-level RGB-D scenes and 1M 2D image scenes to facilitate large-scale POMA-3D pretraining. Experiments show that POMA-3D serves as a strong backbone for both specialist and generalist 3D understanding. It benefits diverse tasks, including 3D question answering, embodied navigation, scene retrieval, and embodied localization, all achieved using only geometric inputs (i.e., 3D coordinates). Overall, our POMA-3D explores a point map way to 3D scene understanding, addressing the scarcity of pretrained priors and limited data in 3D representation learning. Project Page: https://matchlab-imperial.github.io/poma3d/

Junpeng Jing, Weixun Luo, Ye Mao et al.

Recent advances in stereo matching have focused on accuracy, often at the cost of significantly increased model size. Traditionally, the community has regarded efficient models as incapable of zero-shot ability due to their limited capacity. In this paper, we introduce Lite Any Stereo, a stereo depth estimation framework that achieves strong zero-shot generalization while remaining highly efficient. To this end, we design a compact yet expressive backbone to ensure scalability, along with a carefully crafted hybrid cost aggregation module. We further propose a three-stage training strategy on million-scale data to effectively bridge the sim-to-real gap. Together, these components demonstrate that an ultra-light model can deliver strong generalization, ranking 1st across four widely used real-world benchmarks. Remarkably, our model attains accuracy comparable to or exceeding state-of-the-art non-prior-based accurate methods while requiring less than 1% computational cost, setting a new standard for efficient stereo matching.