Sync-NeRF: Generalizing Dynamic NeRFs to Unsynchronized Videos

As a powerful representation of 3D scenes, the neural radiance field (NeRF) enables high-quality novel view synthesis from multi-view images. Stylizing NeRF, however, remains challenging, especially in simulating a text-guided style with both the appearance and the geometry altered simultaneously. In this paper, we present NeRF-Art, a text-guided NeRF stylization approach that manipulates the style of a pre-trained NeRF model with a simple text prompt. Unlike previous approaches that either lack sufficient geometry deformations and texture details or require meshes to guide the stylization, our method can shift a 3D scene to the target style characterized by desired geometry and appearance variations without any mesh guidance. This is achieved by introducing a novel global-local contrastive learning strategy, combined with the directional constraint to simultaneously control both the trajectory and the strength of the target style. Moreover, we adopt a weight regularization method to effectively suppress cloudy artifacts and geometry noises which arise easily when the density field is transformed during geometry stylization. Through extensive experiments on various styles, we demonstrate that our method is effective and robust regarding both single-view stylization quality and cross-view consistency.

Designing a 3D representation of a dynamic scene for fast optimization and rendering is a challenging task. While recent explicit representations enable fast learning and rendering of dynamic radiance fields, they require a dense set of input viewpoints. In this work, we focus on learning a fast representation for dynamic radiance fields with sparse input viewpoints. However, the optimization with sparse input is under-constrained and necessitates the use of motion priors to constrain the learning. Existing fast dynamic scene models do not explicitly model the motion, making them difficult to be constrained with motion priors. We design an explicit motion model as a factorized 4D representation that is fast and can exploit the spatio-temporal correlation of the motion field. We then introduce reliable flow priors including a combination of sparse flow priors across cameras and dense flow priors within cameras to regularize our motion model. Our model is fast, compact and achieves very good performance on popular multi-view dynamic scene datasets with sparse input viewpoints. The source code for our model can be found on our project page: https://nagabhushansn95.github.io/publications/2024/RF-DeRF.html.

Lens arrays enable us to easily achieve inexpensive light field acquisition in comparison with massive camera arrays. However, their micro-lenses often cause insufficient exposure to obtain high quality multi-view videos like camera arrays. In this paper, we propose a novel virtual camera array acquiring dynamic light fields as multi-view videos robustly and inexpensively, where mirror arrays consisting of tangent planes to a paraboloid are effectively combined with a single video camera. First, we demonstrate multi-view imaging of not an object but ordinary scenes with sufficient exposure by using a simple horizontal mirror array implemented as a prototype. Next, we theoretically reveal that there exists smooth extension of our mirror array to the vertical direction, that can be implemented easily by 3D printing, CNC machines, and so on. Finally, our design of light field acquisition is improved for effectively utilizing the whole of its mirror array by introducing an one-way mirror to prevent the camera itself from appearing on multi-view images.

Recent advances in radiance fields enable photorealistic rendering of static or dynamic 3D scenes, but still do not support explicit deformation that is used for scene manipulation or animation. In this paper, we propose a method that enables a new type of deformation of the radiance field: free-form radiance field deformation. We use a triangular mesh that encloses the foreground object called cage as an interface, and by manipulating the cage vertices, our approach enables the free-form deformation of the radiance field. The core of our approach is cage-based deformation which is commonly used in mesh deformation. We propose a novel formulation to extend it to the radiance field, which maps the position and the view direction of the sampling points from the deformed space to the canonical space, thus enabling the rendering of the deformed scene. The deformation results of the synthetic datasets and the real-world datasets demonstrate the effectiveness of our approach. Project page: https://xth430.github.io/deforming-nerf/.KeywordsScene representationRadiance fieldScene manipulationCage-based deformationFree-form deformation

Multi-focus image fusion is an effective way to overcome the shallow depth of field of ordinary cameras and to achieve all-in-focus imaging across the entire depth range. However, existing multi-focus image fusion methods are usually designed for single-view and static scenes, mainly because the current focal stack imaging technology for capturing multi-focus images can only capture one partially focused image at each moment. To simultaneously capture multi-focus images for all-in-focus imaging in dynamic scenes, we propose a computational imaging scheme for all-in-focus dynamic image fusion. Specifically, a multi-view multi-focus (MVMF) imaging technology that integrates the characteristics of both multi-view imaging and focal stack imaging is designed to simultaneously capture MVMF images. For the fusion of MVMF images, we design an MVMF image fusion method based on focus clarity evaluation. This method consists of three modules: image alignment, focus clarity evaluation, and image fusion, which can effectively fuse MVMF images into a high-quality all-in-focus image. In addition, to fill the gap of lacking dynamic multi-focus image datasets in the existing multi-focus image fusion field, we render an MVMF image dataset (i.e., DYN-MVMF) capturing dynamic scenes using the Blender software. On the DYN-MVMF dataset, we conduct extensive comparative experiments with state-of-the-art multi-focus image fusion methods. The experimental results validate the effectiveness of the proposed method in fusing MVMF images into a high-quality all-in-focus image. Moreover, we also conduct comparative experiments with the conventional focal stack imaging scheme on dynamic scene imaging and salient object detection tasks. The results demonstrate that our proposed computational imaging scheme can achieve all-in-focus imaging for each frame of dynamic scenes, thereby enabling more accurate salient object detection results. Therefore, the proposed scheme is expected to contribute to the development of high-performance imaging systems.

We present a novel semantic model for human head defined with neural radiance field. The 3D-consistent head model consist of a set of disentangled and interpretable bases, and can be driven by low-dimensional expression coefficients. Thanks to the powerful representation ability of neural radiance field, the constructed model can represent complex facial attributes including hair, wearings, which can not be represented by traditional mesh blendshape. To construct the personalized semantic facial model, we propose to define the bases as several multi-level voxel fields. With a short monocular RGB video as input, our method can construct the subject's semantic facial NeRF model with only ten to twenty minutes, and can render a photorealistic human head image in tens of miliseconds with a given expression coefficient and view direction. With this novel representation, we apply it to many tasks like facial retargeting and expression editing. Experimental results demonstrate its strong representation ability and training/inference speed. Demo videos and released code are provided in our project page: https://ustc3dv.github.io/NeRFBlendShape/

Land remote-sensing analysis is a crucial research in earth science. In this work, we focus on a challenging task of land analysis, i.e., automatic extraction of traffic roads from remote-sensing data, which has widespread applications in urban development and expansion estimation. Nevertheless, conventional methods either only utilized the limited information of aerial images, or simply fused multimodal information (e.g., vehicle trajectories), thus cannot well recognize unconstrained roads. To facilitate this problem, we introduce a novel neural network framework termed cross-modal message propagation network (CMMPNet), which fully benefits the complementary different modal data (i.e., aerial images and crowdsourced trajectories). Specifically, CMMPNet is composed of two deep autoencoders for modality-specific representation learning and a tailor-designed dual enhancement module for cross-modal representation refinement. In particular, the complementary information of each modality is comprehensively extracted and dynamically propagated to enhance the representation of another modality. Extensive experiments on three real-world benchmarks demonstrate the effectiveness of our CMMPNet for robust road extraction benefiting from blending different modal data, either using image and trajectory data or image and light detection and ranging (LiDAR) data. From the experimental results, we observe that the proposed approach outperforms current state-of-the-art methods by large margins. Our source code is resealed on the project page http://lingboliu.com/multimodal_road_extraction.html.

Novel view synthesis from a single image has recently attracted a lot of attention, and it has been primarily advanced by 3D deep learning and rendering techniques. However, most work is still limited by synthesizing new views within relatively small camera motions. In this paper, we propose a novel approach to synthesize a consistent long-term video given a single scene image and a trajectory of large camera motions. Our approach utilizes an autoregressive Transformer to perform sequential modeling of multiple frames, which reasons the relations between multiple frames and the corresponding cameras to predict the next frame. To facilitate learning and ensure consistency among generated frames, we introduce a locality constraint based on the input cameras to guide self-attention among a large number of patches across space and time. Our method outperforms state-of-the-art view synthesis approaches by a large margin, especially when synthesizing long-term future in indoor 3D scenes. Project page at https://xrenaa.github.io/look-outside-room/.

Visually exploring in a real-world 4D spatiotemporal space freely in VR has been a long-term quest. The task is especially appealing when only a few or even single RGB cameras are used for capturing the dynamic scene. To this end, we present an efficient framework capable of fast reconstruction, compact modeling, and streamable rendering. First, we propose to decompose the 4D spatiotemporal space according to temporal characteristics. Points in the 4D space are associated with probabilities of belonging to three categories: static, deforming, and new areas. Each area is represented and regularized by a separate neural field. Second, we propose a hybrid representations based feature streaming scheme for efficiently modeling the neural fields. Our approach, coined NeRFPlayer, is evaluated on dynamic scenes captured by single hand-held cameras and multi-camera arrays, achieving comparable or superior rendering performance in terms of quality and speed comparable to recent state-of-the-art methods, achieving reconstruction in 10 seconds per frame and interactive rendering. Project website: https://bit.ly/nerfplayer.

Neural rendering techniques aim to generate photorealistic images and accurate 3D geometries from multi-view images but often struggle with efficiency and geometric consistency in complex or dynamic scenes. Optimized Neural Radiance Fields (OptiNeRF) addresses these challenges through several innovations. It uses spatially optimized sampling to focus on points near object surfaces, reducing computation while improving precision. Leveraging the pre-trained Marigold model, it generates depth and normal maps as geometric priors. Sampled points are processed through a hybrid network combining an MLP and a multi-resolution feature grid (MRF), capturing fine details and large-scale structures. To handle varying illumination and complex materials, OptiNeRF introduces adaptive volume rendering (AVR), dynamically adjusting light transparency and scattering. A progressive sampling strategy further focuses computation on regions with high geometric complexity. The loss function incorporates RGB, normal, depth, boundary, and lighting optimization losses, with adaptive weight modulation for geometric priors, ensuring both visual fidelity and geometric consistency even with inaccurate depth/normal estimates. Experiments on dynamic scenes show strong performance, with a PSNR of 32.10 dB, SSIM of 0.936, Chamfer distance of 1.28×10−3, training time of 12 h, and rendering speed of 25 FPS, demonstrating high geometric accuracy, realistic rendering, and computational efficiency over conventional methods.

Representing and synthesizing novel views in real-world dynamic scenes from casual monocular videos is a long-standing problem. Existing solutions typically approach dynamic scenes by applying geometry techniques or utilizing temporal information between several adjacent frames without considering the underlying background distribution in the entire scene or the transmittance over the ray dimension, limiting their performance on static and occlusion areas. Our approach backdrop-driven neural radiance fields offers high-quality view synthesis and a 3D solution to detach the background from the entire dynamic scene, which is called DetRF. Specifically, it employs a neural representation to capture the scene distribution in the static background and a 6D-input NeRF to represent dynamic objects, respectively. Each ray sample is given an additional occlusion weight to indicate the transmittance lying in the static and dynamic components. We evaluate DetRF on public dynamic scenes and our urban driving scenes acquired from an autonomous-driving dataset. Extensive experiments demonstrate that our approach outperforms previous methods in rendering texture details and motion areas while also producing a clean static background. Code will be available soon.

Neural implicit 3D representations have emerged as a powerful paradigm for reconstructing surfaces from multi-view images and synthesizing novel views. Unfortunately, existing methods such as DVR or IDR require accurate per-pixel object masks as supervision. At the same time, neural radiance fields have revolutionized novel view synthesis. However, NeRF’s estimated volume density does not admit accurate surface reconstruction. Our key insight is that implicit surface models and radiance fields can be formulated in a unified way, enabling both surface and volume rendering using the same model. This unified perspective enables novel, more efficient sampling procedures and the ability to reconstruct accurate surfaces without input masks. We compare our method on the DTU, BlendedMVS, and a synthetic indoor dataset. Our experiments demonstrate that we outperform NeRF in terms of reconstruction quality while performing on par with IDR without requiring masks.

This paper proposes a do-it-all neural model of human hands, named LISA. The model can capture accurate hand shape and appearance, generalize to arbitrary hand sub-jects, provide dense surface correspondences, be reconstructed from images in the wild, and can be easily an-imated. We train LISA by minimizing the shape and appearance losses on a large set of multi-view RGB image se-quences annotated with coarse 3D poses of the hand skele-ton. For a 3D point in the local hand coordinates, our model predicts the color and the signed distance with respect to each hand bone independently, and then combines the per-bone predictions using the predicted skinning weights. The shape, color, and pose representations are disentangled by design, enabling fine control of the selected hand param-eters. We experimentally demonstrate that LISA can ac-curately reconstruct a dynamic hand from monocular or multi-view sequences, achieving a noticeably higher qual-ity of reconstructed hand shapes compared to baseline approaches. Project page: https://www.iri.upc.edu/people/ecorona/lisa/.

Multi-beam LiDAR sensors, as used on autonomous vehicles and mobile robots, acquire sequences of 3D range scans (“frames”). Each frame covers the scene sparsely, due to limited angular scanning resolution and occlusion. The sparsity restricts the performance of downstream processes like semantic segmentation or surface reconstruction. Luckily, when the sensor moves, frames are captured from a sequence of different viewpoints. This provides complementary information and, when accumulated in a common scene coordinate frame, yields a denser sampling and a more complete coverage of the underlying 3D scene. However, often the scanned scenes contain moving objects. Points on those objects are not correctly aligned by just undoing the scanner’s ego-motion. In the present paper, we explore multi-frame point cloud accumulation as a mid-level representation of 3D scan sequences, and develop a method that exploits inductive biases of outdoor street scenes, including their geometric layout and object-level rigidity. Compared to state-of-the-art scene flow estimators, our proposed approach aims to align all 3D points in a common reference frame correctly accumulating the points on the individual objects. Our approach greatly reduces the alignment errors on several benchmark datasets. Moreover, the accumulated point clouds benefit high-level tasks like surface reconstruction. [project page]

Generating dynamic scenes from images has gained increasing attention. Existing methods have two major limitations: 1) they can hardly handle sparse images which exhibit limited geometry constraints and insufficient motion; 2) they struggle to maintain spatial-temporal consistency when rendering multi-view videos. To address these limitations, we propose SCSV, a spatial-temporal consistent dynamic scene generation method from sparse views. Our method consists of two stages: scene reconstruction and scene expansion, both of which decouple background and foreground. In the scene reconstruction stage, we first interpolate a set of images between the input images based on a video generation model, followed by the optimization of the scene Gaussian from the interpolated and input images. To improve the spatial-temporal consistency of the reconstructed scene, we propose an uncertainty-aware Gaussian training approach, which introduces adaptive weights of images and pixels. In the scene expansion stage, for background, we render novel views and refine them with a geometry-aware diffusion process. These refined images are then used to incrementally add the Gaussians. As to foreground, we generate human motion according to previous motion, enabling temporal coherent generation of motion. To further enhance the physical plausibility, we integrate the expanded foreground into the background using a gravity-aware alignment. Experiments on NeuMan, Bonn, and EMDB datasets demonstrate that our SCSV achieves superior performance compared to state-of-the-art methods. The code will be released upon acceptance.