CVPR 2025 Papers

Training-free Neural Architecture Search (NAS) efficiently identifies high-performing neural networks using zero-cost (ZC) proxies. Unlike multi-shot and one-shot NAS approaches, ZC-NAS is both (i) time-efficient, eliminating the need for model training, and (ii) interpretable, with proxy designs often theoretically grounded. Despite rapid developments in the field, current SOTA ZC proxies are typically constrained to well-established convolutional search spaces. With the rise of Large Language Models shaping the future of deep learning, this work extends ZC proxy applicability to Vision Transformers (ViTs). We present a new benchmark using the Autoformer search space evaluated on 6 distinct tasks, and propose Layer-Sample Wise Activation with Gradients information (L-SWAG), a novel, generalizable metric that characterises both convolutional and transformer architectures across 14 tasks. Additionally, previous works highlighted how different proxies contain complementary information, motivating the need for a ML model to identify useful combinations. To further enhance ZC-NAS, we therefore introduce LIBRA-NAS (Low Information gain and Bias Re-Alignment), a method that strategically combines proxies to best represent a specific benchmark. Integrated into the NAS search, LIBRA-NAS outperforms evolution and gradient-based NAS techniques by identifying an architecture with a 17.0% test error on ImageNet1k in just 0.1 GPU days.

Neural Radiance Fields (NeRF) has demonstrated its superior capability to represent 3D geometry but require accurately precomputed camera poses during training. To mitigate this requirement, existing methods jointly optimize camera poses and NeRF often relying on good pose initialisation or depth priors. However, these approaches struggle in challenging scenarios, such as large rotations, as they map each camera to a world coordinate system. We propose a novel method that eliminates prior dependencies by modeling continuous camera motions as time-dependent angular velocity and velocity. Relative motions between cameras are learned first via velocity integration, while camera poses can be obtained by aggregating such relative motions up to a world coordinate system defined at a single time step within the video. Specifically, accurate continuous camera movements are learned through a time-dependent NeRF, which captures local scene geometry and motion by training from neighboring frames for each time step. The learned motions enable fine-tuning the NeRF to represent the full scene geometry. Experiments on Co3D and Scannet show our approach achieves superior camera pose and depth estimation and comparable novel-view synthesis performance compared to state-of-the-art methods.

We introduce InteractVLM, a novel method to estimate 3D contact points on human bodies and objects from single in-the-wild images, enabling accurate human-object joint reconstruction in 3D. This is challenging due to occlusions, depth ambiguities, and widely varying object shapes. Existing methods rely on 3D contact annotations collected via expensive motion-capture systems or tedious manual labeling, limiting scalability and generalization. To overcome this, InteractVLM harnesses the broad visual knowledge of large Vision-Language Models (VLMs), fine-tuned with limited 3D contact data. However, directly applying these models is non-trivial, as they reason only in 2D, while human-object contact is inherently 3D. Thus we introduce a novel Render-Localize-Lift module that: (1) embeds 3D body and object surfaces in 2D space via multi-view rendering, (2) trains a novel multi-view localization model (MV-Loc) to infer contacts in 2D, and (3) lifts these to 3D. Additionally, we propose a new task called Semantic Human Contact estimation, where human contact predictions are conditioned explicitly on object semantics, enabling richer interaction modeling. InteractVLM outperforms existing work on contact estimation and also facilitates 3D reconstruction from an in-the wild image.

We propose a generative technique to edit 3D shapes, represented as meshes, NeRFs, or Gaussian Splats, in ~3 seconds, without the need for running an SDS type of optimization.Our key insight is to cast 3D editing as a multiview image inpainting problem, as this representation is generic and can be mapped back to any 3D representation using the bank of available Large Reconstruction Models. We explore different fine-tuning strategies to obtain both multiview generation and inpainting capabilities within the same diffusion model. In particular, the design of the inpainting mask is an important factor of training an inpainting model, and we propose several masking strategies to mimic the types of edits a user would perform on a 3D shape. Our approach takes 3D generative editing from hours to seconds and produces higher-quality results compared to previous works.

Building Free-Viewpoint Videos in a streaming manner offers the advantage of rapid responsiveness compared to offline training methods, greatly enhancing user experience. However, current streaming approaches face challenges of high per-frame reconstruction time (10s+) and error accumulation, limiting their broader application. In this paper, we propose Instant Gaussian Stream (IGS), a fast and generalizable streaming framework, to address these issues. First, we introduce a generalized Anchor-driven Gaussian Motion Network, which projects multi-view 2D motion features into 3D space, using anchor points to drive the motion of all Gaussians. This generalized Network generates the motion of Gaussians for each target frame in the time required for a single inference. Second, we propose a Key-frame-guided Streaming Strategy that refines each key frame, enabling accurate reconstruction of temporally complex scenes while mitigating error accumulation. We conducted extensive in-domain and cross-domain evaluations, demonstrating that our approach can achieve streaming with a average per-frame reconstruction time of 2s+, alongside a enhancement in view synthesis quality.

3D scene understanding is vital for applications in autonomous driving, robotics, and augmented reality. However, scene understanding based on 3D Gaussian Splatting faces three key challenges: (i) an imbalance between appearance and semantics, (ii) inconsistencies in object boundaries, and (iii) difficulties with top-down instance segmentation. To address these challenges, we propose InstanceGaussian, a method that jointly learns appearance and semantic features while adaptively aggregating instances. Our contributions are as follows: (i) a new Semantic-Scaffold-GS representation to improve feature representation and boundary delineation, (ii) a progressive training strategy for enhanced stability and segmentation, and (iii) a category-agnostic, bottom-up instance aggregation approach for better segmentation. Experimental results demonstrate that our approach achieves state-of-the-art performance in category-agnostic, open-vocabulary 3D point-level segmentation, validating the effectiveness of our proposed method.

3D Gaussian Splatting (3DGS) has shown impressive performance in scene reconstruction, offering high rendering quality and rapid rendering speed with short training time. However, it often yields unsatisfactory results when applied to indoor scenes due to its poor ability to learn geometries without enough textural information. In this paper, we propose a new 3DGS-based method "IndoorGS", that leverages the commonly found yet important geometric cues in indoor scenes to improve the reconstruction quality. Specifically, we first extract 2D lines from input images and fuse them into 3D line cues via feature-based matching, which can provide a structural understanding of the target scene. We then apply the statistical outlier removal to refine Structure-from-Motion (SfM) points, ensuring robust cues in texture-rich areas. Based on these two types of cues, we further extract reliable 3D plane cues for textureless regions. Such geometric information will be utilized not only for initialization but also in the realization of a geometric-cue-guided adaptive density control (ADC) strategy. The proposed ADC approach is grounded in the principle of divide-and-conquer and optimizes the use of each type of geometric cues to enhance overall reconstruction performance. Extensive experiments on multiple indoor datasets show that our method can deliver much more accurate geometry and higher rendering quality for indoor scenes than existing 3DGS approaches.

Implicit neural representation and explicit 3D Gaussian Splatting (3D-GS) for novel view synthesis have achieved remarkable progress with frame-based camera (e.g. RGB and RGB-D cameras) recently. Compared to frame-based camera, a novel type of bio-inspired visual sensor, i.e. event camera, has demonstrated advantages in high temporal resolution, high dynamic range, low power consumption and low latency, which make it being favored for many robotic applications. In this work, we present IncEventGS, an incremental 3D Gaussian Splatting reconstruction algorithm with a single event camera, without the assumption of known camera poses. To recover the 3D scene representation incrementally, we exploit the tracking and mapping paradigm of conventional SLAM pipelines for IncEventGS. Given the incoming event stream, the tracker first estimates an initial camera motion based on prior reconstructed 3D-GS scene representation. The mapper then jointly refines both the 3D scene representation and camera motion based on the previously estimated motion trajectory from the tracker. The experimental results demonstrate that IncEventGS delivers superior performance compared to prior NeRF-based methods and other related baselines, even we do not have the ground-truth camera poses. Furthermore, our method can also deliver better performance compared to state-of-the-art event visual odometry methods in terms of camera motion estimation.

In this paper, we observe that the collaboration of various foundation models can perceive semantic and degraded information within images, thereby guiding the low-light enhancement process. Specifically, we propose a self-supervised low-light enhancement framework based on the multiple foundation models collaboration (dubbed FoCo), aimed at improving both the visual quality of enhanced images and the performance in high-level applications. At the feature level, FoCo leverages the rich features from various foundation models to enhance the model's semantic perception during training, thereby reducing the gap between enhanced results and high-quality images from a high-level perspective. At the task level, we exploit the robustness-gap between strong foundation models and weak models, applying high-level task guidance to the low-light enhancement training process. Through the collaboration of multiple foundation models, the proposed framework shows better enhancement performance and adapts better to high-level tasks. Extensive experiments across various enhancement and application benchmarks demonstrate the qualitative and quantitative superiority of the proposed method over numerous state-of-the-art techniques.

Point management is critical for optimizing 3D Gaussian Splatting models, as point initiation (e.g., via structure from motion) is often distributionally inappropriate. Typically, Adaptive Density Control (ADC) algorithm is adopted, leveraging view-averaged gradient magnitude thresholding for point densification, opacity thresholding for pruning, and regular all-points opacity reset. We reveal that this strategy is limited in tackling intricate/special image regions (e.g., transparent) due to inability of identifying all 3D zones requiring point densification, and lacking an appropriate mechanism to handle ill-conditioned points with negative impacts (e.g., occlusion due to false high opacity).To address these limitations, we propose a Localized Point Management(LPM) strategy, capable of identifying those error-contributing zones in greatest need for both point addition and geometry calibration. Zone identification is achieved by leveraging the underlying multiview geometry constraints, subject to image rendering errors. We apply point densification in the identified zones and then reset the opacity of the points in front of these regions, creating a new opportunity to correct poorly conditioned points. Serving as a versatile plugin, LPM can be seamlessly integrated into existing static 3D and dynamic 4D Gaussian Splatting models with minimal additional cost.Experimental evaluations validate the efficacy of our LPM in boosting a variety of existing 3D/4D models both quantitatively and qualitatively. Notably, LPM improves both static 3DGS and dynamic SpaceTimeGS to achieve state-of-the-art rendering quality while retaining real-time speeds, excelling on challenging datasets such as Tanks & Temples and the Neural 3D Video dataset.

Image-to-point cloud registration aims to estimate the camera pose of a given image within a 3D scene point cloud. In this area, matching-based methods have achieved leading performance by first detecting the overlapping region, then matching point and pixel features learned by neural networks and finally using the PnP-RANSAC algorithm to estimate camera pose. However, achieving accurate image-to-point cloud registration remains challenging because the overlapping region detection is unreliable merely relying on point-wise classification, direct alignment of cross-modal data is difficult and indirect optimization objective leads to unstable registration results. To address these challenges, we propose a novel implicit correspondence learning method, including a Geometric Prior-guided overlapping region Detection Module (GPDM), an Implicit Correspondence Learning Module (ICLM), and a Pose Regression Module (PRM). The proposed method enjoys several merits. First, the proposed GPDM can precisely detect the overlapping region. Second, the ICLM can generate robust cross-modality correspondences. Third, the PRM can enable end-to-end optimization. Extensive experimental results on KITTI and nuScenes datasets demonstrate that the proposed model sets a new state-of-the-art performance in registration accuracy.

Score distillation sampling (SDS) demonstrates a powerful capability for text-conditioned 2D image and 3D object generation by distilling the knowledge from learned score functions. However, SDS often suffers from blurriness caused by noisy gradients. When SDS meets the image editing, such degradations can be reduced by adjusting bias shifts using reference pairs, but the de-biasing techniques are still corrupted by erroneous gradients. To this end, we introduce Identity-preserving Distillation Sampling (IDS), which compensates for the gradient leading to undesired changes in the results. Based on the analysis that these errors come from the text-conditioned scores, a new regularization technique, called fixed-point iterative regularization (FPR), is proposed to modify the score itself, driving the preservation of the identity even including poses and structures. Thanks to a self-correction by FPR, the proposed method provides clear and unambiguous representations corresponding to the given prompts in image-to-image editing and editable neural radiance field (NeRF). The structural consistency between the source and the edited data is obviously maintained compared to other state-of-the-art methods.

Inverse rendering seeks to recover 3D geometry, surface material, and lighting from captured images, enabling advanced applications such as novel-view synthesis, relighting, and virtual object insertion. However, most existing techniques rely on high dynamic range (HDR) images as input, limiting accessibility for general users. In response, we introduce IRIS, an inverse rendering framework that recovers the physically based material, spatially-varying HDR lighting, and camera response functions from multi-view, low-dynamic-range (LDR) images. By eliminating the dependence on HDR input, we make inverse rendering technology more accessible. We evaluate our approach on real-world and synthetic scenes and compare it with state-of-the-art methods. Our results show that IRIS effectively recovers HDR lighting, accurate material, and plausible camera response functions, supporting photorealistic relighting and object insertion.

We propose a novel 3D-aware diffusion-based method for generating photorealistic talking head videos directly from a single identity image and explicit control signals (e.g., expressions). Our method generates Multiplane Images (MPIs) that ensure geometric consistency, making them ideal for immersive viewing experiences like binocular videos for VR headsets.Unlike existing methods that often require a separate stage or joint optimization to reconstruct a 3D representation (such as NeRF or 3D Gaussians), our approach directly generates the final output through a single denoising process, eliminating the need for post-processing steps to render novel views efficiently.To effectively learn from monocular videos, we introduce a training mechanism that reconstructs the output MPI randomly in either the target or the reference camera space. This approach enables the model to simultaneously learn sharp image details and underlying 3D information.Extensive experiments demonstrate the effectiveness of our method, which achieves competitive avatar quality and novel-view rendering capabilities, even without explicit 3D reconstruction or high-quality multi-view training data.

This work presents IAAO, a novel framework that builds an explicit 3D model for intelligent agents to gain understanding of articulated objects in their environment through interaction. Unlike prior methods that rely on task-specific networks and assumptions about movable parts, our IAAO leverages large foundation models to estimate interactive affordances and part articulations in three stages. We first build hierarchical features and label fields for each object state using 3D Gaussian Splatting (3DGS) by distilling mask features and view-consistent labels from multi-view images. We then perform object- and part-level queries on the 3D Gaussian primitives to identify static and articulated elements, estimating global transformations and local articulation parameters along with affordances. Finally, scenes from different states are merged and refined based on the estimated transformations, enabling robust affordance-based interaction and manipulation of objects. Experimental results demonstrate the effectiveness of our method.

Recent advancements in computer vision have highlighted the scalability of Vision Transformers (ViTs) across various tasks, yet challenges remain in balancing adaptability, computational efficiency, and the ability to model higher-order relationships. Vision Graph Neural Networks (ViGs) offer an alternative by leveraging graph-based methodologies but are hindered by the computational bottlenecks of clustering algorithms used for edge generation. To address these issues, we propose the Hypergraph Vision Transformer (HgVT), which incorporates a hierarchical bipartite hypergraph structure into the vision transformer framework to capture higher-order semantic relationships while maintaining computational efficiency. HgVT leverages population and diversity regularization for dynamic hypergraph construction without clustering, and expert edge pooling to enhance semantic extraction and facilitate graph-based image retrieval. Empirical results demonstrate that HgVT achieves strong performance on image classification and retrieval, positioning it as an efficient framework for semantic-based vision tasks.

3D point cloud segmentation is essential across a range of applications; however, conventional methods often struggle in evolving environments, particularly when tasked with identifying novel categories under limited supervision. Few-Shot Learning (FSL) and Class Incremental Learning (CIL) have been adapted previously to address these challenges in isolation, yet the combined paradigm of Few-Shot Class Incremental Learning (FSCIL) remains largely unexplored for point cloud segmentation. To address this gap, we introduce Hyperbolic Ideal Prototypes Optimization (HIPO), a novel framework that harnesses hyperbolic embeddings for FSCIL in 3D point clouds. HIPO employs the Poincare Hyperbolic Sphere as its embedding space, integrating Ideal Prototypes enriched by CLIP-derived class semantics, to capture the hierarchical structure of 3D data. By enforcing orthogonality among prototypes and maximizing representational margins, HIPO constructs a resilient embedding space that mitigates forgetting and enables the seamless integration of new classes, thereby effectively countering overfitting. Extensive evaluations on S3DIS, ScanNetv2, and cross-dataset scenarios demonstrate HIPO's strong performance, significantly surpassing existing approaches in both in-domain and cross-dataset FSCIL tasks for 3D point cloud segmentation.

We introduce HyperGS, a novel framework for Hyperspectral Novel View Synthesis (HNVS), based on a new latent 3D Gaussian Splatting (3DGS) technique. Our approach enables simultaneous spatial and spectral renderings by encoding material properties from multi-view 3D hyperspectral datasets. HyperGS reconstructs high-fidelity views from arbitrary perspectives with improved accuracy and speed, outperforming currently existing methods. To address the challenges of high-dimensional data, we perform view synthesis in a learned latent space, incorporating a pixel-wise adaptive density function and a pruning technique for increased training stability and efficiency. Additionally, we introduce the first HNVS benchmark, implementing a number of new baselines based on recent SOTA RGB-NVS techniques, alongside the small number of prior works on HNVS. We demonstrate HyperGS's robustness through extensive evaluation of real and simulated hyperspectral scenes with a 14dB accuracy improvement upon previously published models.