Jun Gao

Jun Gao, Tianchang Shen, Zian Wang et al. · nvidia, utoronto

As several industries are moving towards modeling massive 3D virtual worlds, the need for content creation tools that can scale in terms of the quantity, quality, and diversity of 3D content is becoming evident. In our work, we aim to train performant 3D generative models that synthesize textured meshes which can be directly consumed by 3D rendering engines, thus immediately usable in downstream applications. Prior works on 3D generative modeling either lack geometric details, are limited in the mesh topology they can produce, typically do not support textures, or utilize neural renderers in the synthesis process, which makes their use in common 3D software non-trivial. In this work, we introduce GET3D, a Generative model that directly generates Explicit Textured 3D meshes with complex topology, rich geometric details, and high-fidelity textures. We bridge recent success in the differentiable surface modeling, differentiable rendering as well as 2D Generative Adversarial Networks to train our model from 2D image collections. GET3D is able to generate high-quality 3D textured meshes, ranging from cars, chairs, animals, motorbikes and human characters to buildings, achieving significant improvements over previous methods.

Zian Wang, Tianchang Shen, Jun Gao et al. · nvidia, utoronto

Reconstruction and intrinsic decomposition of scenes from captured imagery would enable many applications such as relighting and virtual object insertion. Recent NeRF based methods achieve impressive fidelity of 3D reconstruction, but bake the lighting and shadows into the radiance field, while mesh-based methods that facilitate intrinsic decomposition through differentiable rendering have not yet scaled to the complexity and scale of outdoor scenes. We present a novel inverse rendering framework for large urban scenes capable of jointly reconstructing the scene geometry, spatially-varying materials, and HDR lighting from a set of posed RGB images with optional depth. Specifically, we use a neural field to account for the primary rays, and use an explicit mesh (reconstructed from the underlying neural field) for modeling secondary rays that produce higher-order lighting effects such as cast shadows. By faithfully disentangling complex geometry and materials from lighting effects, our method enables photorealistic relighting with specular and shadow effects on several outdoor datasets. Moreover, it supports physics-based scene manipulations such as virtual object insertion with ray-traced shadow casting.

Gary Leung, Jun Gao, Xiaohui Zeng et al. · utoronto

Existing transformer-based image backbones typically propagate feature information in one direction from lower to higher-levels. This may not be ideal since the localization ability to delineate accurate object boundaries, is most prominent in the lower, high-resolution feature maps, while the semantics that can disambiguate image signals belonging to one object vs. another, typically emerges in a higher level of processing. We present Hierarchical Inter-Level Attention (HILA), an attention-based method that captures Bottom-Up and Top-Down Updates between features of different levels. HILA extends hierarchical vision transformer architectures by adding local connections between features of higher and lower levels to the backbone encoder. In each iteration, we construct a hierarchy by having higher-level features compete for assignments to update lower-level features belonging to them, iteratively resolving object-part relationships. These improved lower-level features are then used to re-update the higher-level features. HILA can be integrated into the majority of hierarchical architectures without requiring any changes to the base model. We add HILA into SegFormer and the Swin Transformer and show notable improvements in accuracy in semantic segmentation with fewer parameters and FLOPS. Project website and code: https://www.cs.toronto.edu/~garyleung/hila/

Katja Schwarz, Seung Wook Kim, Jun Gao et al.

Modern learning-based approaches to 3D-aware image synthesis achieve high photorealism and 3D-consistent viewpoint changes for the generated images. Existing approaches represent instances in a shared canonical space. However, for in-the-wild datasets a shared canonical system can be difficult to define or might not even exist. In this work, we instead model instances in view space, alleviating the need for posed images and learned camera distributions. We find that in this setting, existing GAN-based methods are prone to generating flat geometry and struggle with distribution coverage. We hence propose WildFusion, a new approach to 3D-aware image synthesis based on latent diffusion models (LDMs). We first train an autoencoder that infers a compressed latent representation, which additionally captures the images' underlying 3D structure and enables not only reconstruction but also novel view synthesis. To learn a faithful 3D representation, we leverage cues from monocular depth prediction. Then, we train a diffusion model in the 3D-aware latent space, thereby enabling synthesis of high-quality 3D-consistent image samples, outperforming recent state-of-the-art GAN-based methods. Importantly, our 3D-aware LDM is trained without any direct supervision from multiview images or 3D geometry and does not require posed images or learned pose or camera distributions. It directly learns a 3D representation without relying on canonical camera coordinates. This opens up promising research avenues for scalable 3D-aware image synthesis and 3D content creation from in-the-wild image data. See https://katjaschwarz.github.io/wildfusion for videos of our 3D results.

Chengdong Ma, Ziran Yang, Hai Ci et al.

The primary challenge in deploying Large Language Model (LLM) is ensuring its harmlessness. Red team can identify vulnerabilities by attacking LLM to attain safety. However, current efforts heavily rely on single-round prompt designs and unilateral red team optimizations against fixed blue teams. These static approaches lead to significant reductions in generation diversity, known as the mode collapse, which makes it difficult to discover the potential risks in the increasingly complex human-LLM interactions. Here we introduce dynamic Red Team Game (RTG) to comprehensively analyze the multi-round offensive and defensive interactions between red team and blue team. Furthermore, we develop a Gamified Red Team Solver (GRTS) with diversity measures to mitigate mode collapse and theoretically guarantee the convergence of approximate Nash equilibrium which results in better strategies for both teams. Empirical results demonstrate that GRTS explore diverse and implicit attacks to adaptively exploit various LLMs, surpassing the constraints of specific modes. Insightfully, the geometrical structure we unveil of the red team task aligns with the spinning top hypothesis, confirming the necessity of constructing a diverse LLM population as a promising proxy for heterogeneous human expert red-teamers. This paves the way for scalable toxicity detection and safe alignment for LLMs.

Yu Xie, Qian Qiao, Jun Gao et al.

More and more end-to-end text spotting methods based on Transformer architecture have demonstrated superior performance. These methods utilize a bipartite graph matching algorithm to perform one-to-one optimal matching between predicted objects and actual objects. However, the instability of bipartite graph matching can lead to inconsistent optimization targets, thereby affecting the training performance of the model. Existing literature applies denoising training to solve the problem of bipartite graph matching instability in object detection tasks. Unfortunately, this denoising training method cannot be directly applied to text spotting tasks, as these tasks need to perform irregular shape detection tasks and more complex text recognition tasks than classification. To address this issue, we propose a novel denoising training method (DNTextSpotter) for arbitrary-shaped text spotting. Specifically, we decompose the queries of the denoising part into noised positional queries and noised content queries. We use the four Bezier control points of the Bezier center curve to generate the noised positional queries. For the noised content queries, considering that the output of the text in a fixed positional order is not conducive to aligning position with content, we employ a masked character sliding method to initialize noised content queries, thereby assisting in the alignment of text content and position. To improve the model's perception of the background, we further utilize an additional loss function for background characters classification in the denoising training part.Although DNTextSpotter is conceptually simple, it outperforms the state-of-the-art methods on four benchmarks (Total-Text, SCUT-CTW1500, ICDAR15, and Inverse-Text), especially yielding an improvement of 11.3% against the best approach in Inverse-Text dataset.

Kunyuan Li, Jun Zhang, Rui Sun et al.

Light field cameras record not only the spatial information of observed scenes but also the directions of all incoming light rays. The spatial and angular information implicitly contain geometrical characteristics such as multi-view or epipolar geometry, which can be exploited to improve the performance of depth estimation. An Epipolar Plane Image (EPI), the unique 2D spatial-angular slice of the light field, contains patterns of oriented lines. The slope of these lines is associated with the disparity. Benefiting from this property of EPIs, some representative methods estimate depth maps by analyzing the disparity of each line in EPIs. However, these methods often extract the optimal slope of the lines from EPIs while ignoring the relationship between neighboring pixels, which leads to inaccurate depth map predictions. Based on the observation that an oriented line and its neighboring pixels in an EPI share a similar linear structure, we propose an end-to-end fully convolutional network (FCN) to estimate the depth value of the intersection point on the horizontal and vertical EPIs. Specifically, we present a new feature-extraction module, called Oriented Relation Module (ORM), that constructs the relationship between the line orientations. To facilitate training, we also propose a refocusing-based data augmentation method to obtain different slopes from EPIs of the same scene point. Extensive experiments verify the efficacy of learning relations and show that our approach is competitive to other state-of-the-art methods. The code and the trained models are available at https://github.com/lkyahpu/EPI_ORM.git.

Xuanchi Ren, Tianchang Shen, Jiahui Huang et al. · nvidia, utoronto

We present GEN3C, a generative video model with precise Camera Control and temporal 3D Consistency. Prior video models already generate realistic videos, but they tend to leverage little 3D information, leading to inconsistencies, such as objects popping in and out of existence. Camera control, if implemented at all, is imprecise, because camera parameters are mere inputs to the neural network which must then infer how the video depends on the camera. In contrast, GEN3C is guided by a 3D cache: point clouds obtained by predicting the pixel-wise depth of seed images or previously generated frames. When generating the next frames, GEN3C is conditioned on the 2D renderings of the 3D cache with the new camera trajectory provided by the user. Crucially, this means that GEN3C neither has to remember what it previously generated nor does it have to infer the image structure from the camera pose. The model, instead, can focus all its generative power on previously unobserved regions, as well as advancing the scene state to the next frame. Our results demonstrate more precise camera control than prior work, as well as state-of-the-art results in sparse-view novel view synthesis, even in challenging settings such as driving scenes and monocular dynamic video. Results are best viewed in videos. Check out our webpage! https://research.nvidia.com/labs/toronto-ai/GEN3C/

Yifan Lu, Xuanchi Ren, Jiawei Yang et al. · nvidia, utoronto

We present InfiniCube, a scalable method for generating unbounded dynamic 3D driving scenes with high fidelity and controllability. Previous methods for scene generation either suffer from limited scales or lack geometric and appearance consistency along generated sequences. In contrast, we leverage the recent advancements in scalable 3D representation and video models to achieve large dynamic scene generation that allows flexible controls through HD maps, vehicle bounding boxes, and text descriptions. First, we construct a map-conditioned sparse-voxel-based 3D generative model to unleash its power for unbounded voxel world generation. Then, we re-purpose a video model and ground it on the voxel world through a set of carefully designed pixel-aligned guidance buffers, synthesizing a consistent appearance. Finally, we propose a fast feed-forward approach that employs both voxel and pixel branches to lift the dynamic videos to dynamic 3D Gaussians with controllable objects. Our method can generate controllable and realistic 3D driving scenes, and extensive experiments validate the effectiveness and superiority of our model.

Ruofan Liang, Zan Gojcic, Huan Ling et al. · utoronto

Understanding and modeling lighting effects are fundamental tasks in computer vision and graphics. Classic physically-based rendering (PBR) accurately simulates the light transport, but relies on precise scene representations--explicit 3D geometry, high-quality material properties, and lighting conditions--that are often impractical to obtain in real-world scenarios. Therefore, we introduce DiffusionRenderer, a neural approach that addresses the dual problem of inverse and forward rendering within a holistic framework. Leveraging powerful video diffusion model priors, the inverse rendering model accurately estimates G-buffers from real-world videos, providing an interface for image editing tasks, and training data for the rendering model. Conversely, our rendering model generates photorealistic images from G-buffers without explicit light transport simulation. Experiments demonstrate that DiffusionRenderer effectively approximates inverse and forwards rendering, consistently outperforming the state-of-the-art. Our model enables practical applications from a single video input--including relighting, material editing, and realistic object insertion.

Zekun Jiang, Dongjie Cheng, Ziyuan Qin et al.

This study presents a novel multimodal medical image zero-shot segmentation algorithm named the text-visual-prompt segment anything model (TV-SAM) without any manual annotations. The TV-SAM incorporates and integrates the large language model GPT-4, the vision language model GLIP, and the SAM to autonomously generate descriptive text prompts and visual bounding box prompts from medical images, thereby enhancing the SAM's capability for zero-shot segmentation. Comprehensive evaluations are implemented on seven public datasets encompassing eight imaging modalities to demonstrate that TV-SAM can effectively segment unseen targets across various modalities without additional training. TV-SAM significantly outperforms SAM AUTO and GSAM, closely matching the performance of SAM BBOX with gold standard bounding box prompts and surpasses the state-of-the-art methods on specific datasets such as ISIC and WBC. The study indicates that TV-SAM serves as an effective multimodal medical image zero-shot segmentation algorithm, highlighting the significant contribution of GPT-4 to zero-shot segmentation. By integrating foundational models such as GPT-4, GLIP, and SAM, the ability to address complex problems in specialized domains can be enhanced.

Zhuoyuan Wu, Xurui Yang, Jiahui Huang et al.

Estimating accurate camera poses, 3D scene geometry, and object motion from in-the-wild videos is a long-standing challenge for classical structure from motion pipelines due to the presence of dynamic objects. Recent learning-based methods attempt to overcome this challenge by training motion estimators to filter dynamic objects and focus on the static background. However, their performance is largely limited by the availability of large-scale motion segmentation datasets, resulting in inaccurate segmentation and, therefore, inferior structural 3D understanding. In this work, we introduce the Dynamic Prior (\ourmodel) to robustly identify dynamic objects without task-specific training, leveraging the powerful reasoning capabilities of Vision-Language Models (VLMs) and the fine-grained spatial segmentation capacity of SAM2. \ourmodel can be seamlessly integrated into state-of-the-art pipelines for camera pose optimization, depth reconstruction, and 4D trajectory estimation. Extensive experiments on both synthetic and real-world videos demonstrate that \ourmodel not only achieves state-of-the-art performance on motion segmentation, but also significantly improves accuracy and robustness for structural 3D understanding.

Wanfu Gao, Zengyao Man, Zebin He et al.

Feature generation is a critical step in machine learning, aiming to enhance model performance by capturing complex relationships within the data and generating meaningful new features. Traditional feature generation methods heavily rely on domain expertise and manual intervention, making the process labor-intensive and challenging to adapt to different scenarios. Although automated feature generation techniques address these issues to some extent, they often face challenges such as feature redundancy, inefficiency in feature space exploration, and limited adaptability to diverse datasets and tasks. To address these problems, we propose a Two-Stage Feature Generation (TSFG) framework, which integrates a Transformer-based encoder-decoder architecture with Proximal Policy Optimization (PPO). The encoder-decoder model in TSFG leverages the Transformer's self-attention mechanism to efficiently represent and transform features, capturing complex dependencies within the data. PPO further enhances TSFG by dynamically adjusting the feature generation strategy based on task-specific feedback, optimizing the process for improved performance and adaptability. TSFG dynamically generates high-quality feature sets, significantly improving the predictive performance of machine learning models. Experimental results demonstrate that TSFG outperforms existing state-of-the-art methods in terms of feature quality and adaptability.

Aysegul Dundar, Jun Gao, Andrew Tao et al.

This paper presents a method to reconstruct high-quality textured 3D models from single images. Current methods rely on datasets with expensive annotations; multi-view images and their camera parameters. Our method relies on GAN generated multi-view image datasets which have a negligible annotation cost. However, they are not strictly multi-view consistent and sometimes GANs output distorted images. This results in degraded reconstruction qualities. In this work, to overcome these limitations of generated datasets, we have two main contributions which lead us to achieve state-of-the-art results on challenging objects: 1) A robust multi-stage learning scheme that gradually relies more on the models own predictions when calculating losses, 2) A novel adversarial learning pipeline with online pseudo-ground truth generations to achieve fine details. Our work provides a bridge from 2D supervisions of GAN models to 3D reconstruction models and removes the expensive annotation efforts. We show significant improvements over previous methods whether they were trained on GAN generated multi-view images or on real images with expensive annotations. Please visit our web-page for 3D visuals: https://research.nvidia.com/labs/adlr/progressive-3d-learning

Jacob Munkberg, Jon Hasselgren, Tianchang Shen et al.

We present an efficient method for joint optimization of topology, materials and lighting from multi-view image observations. Unlike recent multi-view reconstruction approaches, which typically produce entangled 3D representations encoded in neural networks, we output triangle meshes with spatially-varying materials and environment lighting that can be deployed in any traditional graphics engine unmodified. We leverage recent work in differentiable rendering, coordinate-based networks to compactly represent volumetric texturing, alongside differentiable marching tetrahedrons to enable gradient-based optimization directly on the surface mesh. Finally, we introduce a differentiable formulation of the split sum approximation of environment lighting to efficiently recover all-frequency lighting. Experiments show our extracted models used in advanced scene editing, material decomposition, and high quality view interpolation, all running at interactive rates in triangle-based renderers (rasterizers and path tracers). Project website: https://nvlabs.github.io/nvdiffrec/ .

Wenzheng Chen, Joey Litalien, Jun Gao et al.

We consider the challenging problem of predicting intrinsic object properties from a single image by exploiting differentiable renderers. Many previous learning-based approaches for inverse graphics adopt rasterization-based renderers and assume naive lighting and material models, which often fail to account for non-Lambertian, specular reflections commonly observed in the wild. In this work, we propose DIBR++, a hybrid differentiable renderer which supports these photorealistic effects by combining rasterization and ray-tracing, taking the advantage of their respective strengths -- speed and realism. Our renderer incorporates environmental lighting and spatially-varying material models to efficiently approximate light transport, either through direct estimation or via spherical basis functions. Compared to more advanced physics-based differentiable renderers leveraging path tracing, DIBR++ is highly performant due to its compact and expressive shading model, which enables easy integration with learning frameworks for geometry, reflectance and lighting prediction from a single image without requiring any ground-truth. We experimentally demonstrate that our approach achieves superior material and lighting disentanglement on synthetic and real data compared to existing rasterization-based approaches and showcase several artistic applications including material editing and relighting.

Yuxuan Zhang, Huan Ling, Jun Gao et al.

We introduce DatasetGAN: an automatic procedure to generate massive datasets of high-quality semantically segmented images requiring minimal human effort. Current deep networks are extremely data-hungry, benefiting from training on large-scale datasets, which are time consuming to annotate. Our method relies on the power of recent GANs to generate realistic images. We show how the GAN latent code can be decoded to produce a semantic segmentation of the image. Training the decoder only needs a few labeled examples to generalize to the rest of the latent space, resulting in an infinite annotated dataset generator! These generated datasets can then be used for training any computer vision architecture just as real datasets are. As only a few images need to be manually segmented, it becomes possible to annotate images in extreme detail and generate datasets with rich object and part segmentations. To showcase the power of our approach, we generated datasets for 7 image segmentation tasks which include pixel-level labels for 34 human face parts, and 32 car parts. Our approach outperforms all semi-supervised baselines significantly and is on par with fully supervised methods, which in some cases require as much as 100x more annotated data as our method.

Yuxuan Zhang, Wenzheng Chen, Huan Ling et al.

Differentiable rendering has paved the way to training neural networks to perform "inverse graphics" tasks such as predicting 3D geometry from monocular photographs. To train high performing models, most of the current approaches rely on multi-view imagery which are not readily available in practice. Recent Generative Adversarial Networks (GANs) that synthesize images, in contrast, seem to acquire 3D knowledge implicitly during training: object viewpoints can be manipulated by simply manipulating the latent codes. However, these latent codes often lack further physical interpretation and thus GANs cannot easily be inverted to perform explicit 3D reasoning. In this paper, we aim to extract and disentangle 3D knowledge learned by generative models by utilizing differentiable renderers. Key to our approach is to exploit GANs as a multi-view data generator to train an inverse graphics network using an off-the-shelf differentiable renderer, and the trained inverse graphics network as a teacher to disentangle the GAN's latent code into interpretable 3D properties. The entire architecture is trained iteratively using cycle consistency losses. We show that our approach significantly outperforms state-of-the-art inverse graphics networks trained on existing datasets, both quantitatively and via user studies. We further showcase the disentangled GAN as a controllable 3D "neural renderer", complementing traditional graphics renderers.

Jun Gao, Zian Wang, Jinchen Xuan et al.

In modern computer vision, images are typically represented as a fixed uniform grid with some stride and processed via a deep convolutional neural network. We argue that deforming the grid to better align with the high-frequency image content is a more effective strategy. We introduce \emph{Deformable Grid} DefGrid, a learnable neural network module that predicts location offsets of vertices of a 2-dimensional triangular grid, such that the edges of the deformed grid align with image boundaries. We showcase our DefGrid in a variety of use cases, i.e., by inserting it as a module at various levels of processing. We utilize DefGrid as an end-to-end \emph{learnable geometric downsampling} layer that replaces standard pooling methods for reducing feature resolution when feeding images into a deep CNN. We show significantly improved results at the same grid resolution compared to using CNNs on uniform grids for the task of semantic segmentation. We also utilize DefGrid at the output layers for the task of object mask annotation, and show that reasoning about object boundaries on our predicted polygonal grid leads to more accurate results over existing pixel-wise and curve-based approaches. We finally showcase DefGrid as a standalone module for unsupervised image partitioning, showing superior performance over existing approaches. Project website: http://www.cs.toronto.edu/~jungao/def-grid

Wenzheng Chen, Jun Gao, Huan Ling et al.

Many machine learning models operate on images, but ignore the fact that images are 2D projections formed by 3D geometry interacting with light, in a process called rendering. Enabling ML models to understand image formation might be key for generalization. However, due to an essential rasterization step involving discrete assignment operations, rendering pipelines are non-differentiable and thus largely inaccessible to gradient-based ML techniques. In this paper, we present {\emph DIB-R}, a differentiable rendering framework which allows gradients to be analytically computed for all pixels in an image. Key to our approach is to view foreground rasterization as a weighted interpolation of local properties and background rasterization as a distance-based aggregation of global geometry. Our approach allows for accurate optimization over vertex positions, colors, normals, light directions and texture coordinates through a variety of lighting models. We showcase our approach in two ML applications: single-image 3D object prediction, and 3D textured object generation, both trained using exclusively using 2D supervision. Our project website is: https://nv-tlabs.github.io/DIB-R/

Jun Gao, Xiao Li, Liwei Wang et al.

We present a method for compositing virtual objects into a photograph such that the object colors appear to have been processed by the photo's camera imaging pipeline. Compositing in such a camera-aware manner is essential for high realism, and it requires the color transformation in the photo's pipeline to be inferred, which is challenging due to the inherent one-to-many mapping that exists from a scene to a photo. To address this problem for the case of a single photo taken from an unknown camera, we propose a dual-learning approach in which the reverse color transformation (from the photo to the scene) is jointly estimated. Learning of the reverse transformation is used to facilitate learning of the forward mapping, by enforcing cycle consistency of the two processes. We additionally employ a feature sharing schema to extract evidence from the target photo in the reverse mapping to guide the forward color transformation. Our dual-learning approach achieves object compositing results that surpass those of alternative techniques.

Huan Ling, Jun Gao, Amlan Kar et al.

Manually labeling objects by tracing their boundaries is a laborious process. In Polygon-RNN++ the authors proposed Polygon-RNN that produces polygonal annotations in a recurrent manner using a CNN-RNN architecture, allowing interactive correction via humans-in-the-loop. We propose a new framework that alleviates the sequential nature of Polygon-RNN, by predicting all vertices simultaneously using a Graph Convolutional Network (GCN). Our model is trained end-to-end. It supports object annotation by either polygons or splines, facilitating labeling efficiency for both line-based and curved objects. We show that Curve-GCN outperforms all existing approaches in automatic mode, including the powerful PSP-DeepLab and is significantly more efficient in interactive mode than Polygon-RNN++. Our model runs at 29.3ms in automatic, and 2.6ms in interactive mode, making it 10x and 100x faster than Polygon-RNN++.

Jun Gao, Chengcheng Tang, Vignesh Ganapathi-Subramanian et al.

Reconstruction of geometry based on different input modes, such as images or point clouds, has been instrumental in the development of computer aided design and computer graphics. Optimal implementations of these applications have traditionally involved the use of spline-based representations at their core. Most such methods attempt to solve optimization problems that minimize an output-target mismatch. However, these optimization techniques require an initialization that is close enough, as they are local methods by nature. We propose a deep learning architecture that adapts to perform spline fitting tasks accordingly, providing complementary results to the aforementioned traditional methods. We showcase the performance of our approach, by reconstructing spline curves and surfaces based on input images or point clouds.