Recent advances in implicit neural representations and differentiable rendering make it possible to simultaneously recover the geometry and materials of an object from multi-view RGB images captured under unknown static illumination. Despite the promising results achieved, indirect illumination is rarely modeled in previous methods, as it requires expensive recursive path tracing which makes the inverse rendering computationally intractable. In this paper, we propose a novel approach to efficiently recovering spatially-varying indirect illumination. The key insight is that indirect illumination can be conveniently derived from the neural radiance field learned from input images instead of being estimated jointly with direct illumination and materials. By properly modeling the indirect illumination and visibility of direct illumination, interreflection- and shadow-free albedo can be recovered. The experiments on both synthetic and real data demonstrate the superior performance of our approach compared to previous work and its capability to synthesize realistic renderings under novel viewpoints and illumination. Our code and data are available at https://zju3dv.github.io/invrender/.
This paper presents an approach to estimating the continuous 6-DoF pose of an object from a single RGB image. The approach combines semantic keypoints predicted by a convolutional network (convnet) with a deformable shape model. Unlike prior investigators, we are agnostic to whether the object is textured or textureless, as the convnet learns the optimal representation from the available training-image data. Furthermore, the approach can be applied to instance- and class-based pose recovery. Additionally, we accompany our main pipeline with a technique for semi-automatic data generation from unlabeled videos. This procedure allows us to train the learnable components of our method with minimal manual intervention in the labeling process. Empirically, we show that our approach can accurately recover the 6-DoF object pose for both instance- and class-based scenarios even against a cluttered background. We apply our approach both to several, existing, large-scale datasets - including PASCAL3D+, LineMOD-Occluded, YCB-Video, and TUD-Light - and, using our labeling pipeline, to a new dataset with novel object classes that we introduce here. Extensive empirical evaluations show that our approach is able to provide pose estimation results comparable to the state of the art.
Large-scale training data with high-quality annotations is critical for training semantic and instance segmentation models. Unfortunately, pixel-wise annotation is labor-intensive and costly, raising the demand for more efficient labeling strategies. In this work, we present a novel 3D-to-2D label transfer method, Panoptic NeRF, which aims for obtaining per-pixel 2D semantic and instance labels from easy-to-obtain coarse 3D bounding primitives. Our method utilizes NeRF as a differentiable tool to unify coarse 3D annotations and 2D semantic cues transferred from existing datasets. We demonstrate that this combination allows for improved geometry guided by semantic information, enabling rendering of accurate semantic maps across multiple views. Furthermore, this fusion process resolves label ambiguity of the coarse 3D annotations and filters noise in the 2D predictions. By inferring in 3D space and rendering to 2D labels, our 2D semantic and instance labels are multi-view consistent by design. Experimental results show that Panoptic NeRF outperforms existing semantic and instance label transfer methods in terms of accuracy and multi-view consistency on challenging urban scenes of the KITTI-360 dataset.
Generating speech-consistent body and gesture movements is a long-standing problem in virtual avatar creation. Previous studies often synthesize pose movement in a holistic manner, where poses of all joints are generated simultaneously. Such a straightforward pipeline fails to generate fine-grained co-speech gestures. One observation is that the hierarchical semantics in speech and the hierarchical structures of human gestures can be naturally described into multiple granularities and associated together. To fully utilize the rich connections between speech audio and human gestures, we propose a novel framework named Hierarchical Audio-to-Gesture (HA2G) for co-speech gesture generation. In HA2G, a Hierarchical Audio Learner extracts audio representations across semantic granularities. A Hierarchical Pose Inferer subsequently renders the entire human pose gradually in a hierarchical manner. To enhance the quality of synthesized gestures, we develop a contrastive learning strategy based on audio-text alignment for better audio representations. Extensive experiments and human evaluation demonstrate that the proposed method renders realistic co-speech gestures and outperforms previous methods in a clear margin. Project page: https://alvinliu0.github.io/projects/HA2G
This paper aims to reconstruct an animatable human model from a video of very sparse camera views. Some recent works represent human geometry and appearance with neural radiance fields and utilize parametric human models to produce deformation fields for animation, which enables them to recover detailed 3D human models from videos. However, their reconstruction results tend to be noisy due to the lack of surface constraints on radiance fields. Moreover, as they generate the human appearance in 3D space, their rendering quality heavily depends on the accuracy of deformation fields. To solve these problems, we propose Animatable Neural Implicit Surface (AniSDF), which models the human geometry with a signed distance field and defers the appearance generation to the 2D image space with a 2D neural renderer. The signed distance field naturally regularizes the learned geometry, enabling the high-quality reconstruction of human bodies, which can be further used to improve the rendering speed. Moreover, the 2D neural renderer can be learned to compensate for geometric errors, making the rendering more robust to inaccurate deformations. Experiments on several datasets show that the proposed approach outperforms recent human reconstruction and synthesis methods by a large margin.
The task of audio-visual sound source localization has been well studied under constrained scenes, where the audio recordings are clean. However, in real-world scenarios, audios are usually contaminated by off-screen sound and background noise. They will interfere with the procedure of identifying desired sources and building visual-sound connections, making previous studies non-applicable. In this work, we propose the Interference Eraser (IEr) framework, which tackles the problem of audio-visual sound source localization in the wild. The key idea is to eliminate the interference by redefining and carving discriminative audio representations. Specifically, we observe that the previous practice of learning only a single audio representation is insufficient due to the additive nature of audio signals. We thus extend the audio representation with our Audio-Instance-Identifier module, which clearly distinguishes sounding instances when audio signals of different volumes are unevenly mixed. Then we erase the influence of the audible but off-screen sounds and the silent but visible objects by a Cross-modal Referrer module with cross-modality distillation. Quantitative and qualitative evaluations demonstrate that our proposed framework achieves superior results on sound localization tasks, especially under real-world scenarios. Code is available at https://github.com/alvinliu0/Visual-Sound-Localization-in-the-Wild.
This paper aims to reduce the rendering time of generalizable radiance fields. Some recent works equip neural radiance fields with image encoders and are able to generalize across scenes, which avoids the per-scene optimization. However, their rendering process is generally very slow. A major factor is that they sample lots of points in empty space when inferring radiance fields. In this paper, we present a hybrid scene representation which combines the best of implicit radiance fields and explicit depth maps for efficient rendering. Specifically, we first build the cascade cost volume to efficiently predict the coarse geometry of the scene. The coarse geometry allows us to sample few points near the scene surface and significantly improves the rendering speed. This process is fully differentiable, enabling us to jointly learn the depth prediction and radiance field networks from only RGB images. Experiments show that the proposed approach exhibits state-of-the-art performance on the DTU, Real Forward-facing and NeRF Synthetic datasets, while being at least 50 times faster than previous generalizable radiance field methods. We also demonstrate the capability of our method to synthesize free-viewpoint videos of dynamic human performers in real-time. The code will be available at https://zju3dv.github.io/enerf/.
Graph neural networks have emerged as a powerful model for graph representation learning to undertake graph-level prediction tasks. Various graph pooling methods have been developed to coarsen an input graph into a succinct graph-level representation through aggregating node embeddings obtained via graph convolution. However, most graph pooling methods are heavily node-centric and are unable to fully leverage the crucial information contained in global graph structure. This paper presents a cross-view graph pooling (Co-Pooling) method to better exploit crucial graph structure information. The proposed Co-Pooling fuses pooled representations learnt from both node view and edge view. Through cross-view interaction, edge-view pooling and node-view pooling seamlessly reinforce each other to learn more informative graph-level representations. Co-Pooling has the advantage of handling various graphs with different types of node attributes. Extensive experiments on a total of 15 graph benchmark datasets validate the effectiveness of our proposed method, demonstrating its superior performance over state-of-the-art pooling methods on both graph classification and graph regression tasks.
We present a new neural representation, called Neural Ray (NeuRay), for the novel view synthesis (NVS) task with multi-view images as input. Existing neural scene representations for solving the NVS problem, such as NeRF, cannot generalize to new scenes and take an excessively long time on training on each new scene from scratch. The other subsequent neural rendering methods based on stereo matching, such as PixelNeRF, SRF and IBRNet are designed to generalize to unseen scenes but suffer from view inconsistency in complex scenes with self-occlusions. To address these issues, our NeuRay method represents every scene by encoding the visibility of rays associated with the input views. This neural representation can efficiently be initialized from depths estimated by external MVS methods, which is able to generalize to new scenes and achieves satisfactory rendering images without any training on the scene. Then, the initialized NeuRay can be further optimized on every scene with little training timing to enforce spatial coherence to ensure view consistency in the presence of severe self-occlusion. Experiments demonstrate that NeuRay can quickly generate high-quality novel view images of unseen scenes with little finetuning and can handle complex scenes with severe self-occlusions which previous methods struggle with.