ROOTS learns to represent and render 3D scenes with object-centric models.
problem Learning to represent and render 3D scenes with object-centric compositionality.
method Probabilistic generative model for learning object representations and scene rendering from partial observations.
result The model can infer 3D object representations and render scenes from arbitrary viewpoints.
3D object recognition accuracy can be improved by learning the multi-scale spatial features from 3D spatial geometric representations of objects such as point clouds, 3D models, surfaces, and RGB-D data. Current deep learning approaches learn such features either using structured data representations (voxel grids and o…
InSphereNet uses infilling spheres for 3D object classification, improving accuracy with fewer parameters.
problem 3D object classification using points, voxels, or images.
method Constructs infilling spheres from signed distance field (SDF) for classification.
result InSphereNet achieves superior accuracy with fewer inputs and parameters.
Paper develops a new method to analyze 3D tree-like objects.
problem Analyzing complex geometrical and topological variations in 3D tree-like objects.
method Extended SRVF representation and new metric for tree-shaped 3D objects.
result Captures full elasticity and topological variations of branches.
New method encodes 3D object geometry into neural network weights for efficient reconstruction.
problem Efficiently representing and reconstructing 3D objects with minimal parameters.
method Mapping network that encodes object geometry into neural network weights, reconstructing objects using simple geometric spaces.
result Reconstructed objects have accuracy comparable to state-of-the-art methods with significantly fewer parameters.
Deep generative architectures provide a way to model not only images but also complex, 3-dimensional objects, such as point clouds. In this work, we present a novel method to obtain meaningful representations of 3D shapes that can be used for challenging tasks including 3D points generation, reconstruction, compression…
Geometric Capsule Autoencoders group 3D points into parts and objects.
problem Learning object representations from 3D point clouds.
method Geometric capsules with pose and feature components, Multi-View Agreement voting mechanism.
result Learned representations enable object identification and canonical pose recovery.
Proposes a model to generate 3D-aware images from 2D images.
problem Generating 3D-aware images from 2D images.
method Likelihood-based top-down model using Neural Radiance Fields and energy-based latent variables.
result Model can infer 3D object structures from 2D images and generate novel views.
Recent progress in deep generative models has led to tremendous breakthroughs in image generation. However, while existing models can synthesize photorealistic images, they lack an understanding of our underlying 3D world. We present a new generative model, Visual Object Networks (VON), synthesizing natural images of o…
The success of various applications including robotics, digital content creation, and visualization demand a structured and abstract representation of the 3D world from limited sensor data. Inspired by the nature of human perception of 3D shapes as a collection of simple parts, we explore such an abstract shape represe…
Generative Multisensory Network learns 3D scene representations from multiple modalities.
problem Learning robust 3D scene representations from multiple sensory modalities.
method Amortized Product-of-Experts for efficient inference and cross-modal generation.
result The model can infer modality-invariant 3D scene representations efficiently from various sensory modalities.
3D Convolutional Neural Networks are sensitive to transformations applied to their input. This is a problem because a voxelized version of a 3D object, and its rotated clone, will look unrelated to each other after passing through to the last layer of a network. Instead, an idealized model would preserve a meaningful r…
3D convolutional neural networks (3D-CNN) have been used for object recognition based on the voxelized shape of an object. In this paper, we present a 3D-CNN based method to learn distinct local geometric features of interest within an object. In this context, the voxelized representation may not be sufficient to captu…
PolyGen models 3D meshes directly, predicting vertices and faces sequentially.
problem Efficiently modeling 3D geometry for computer graphics, robotics, and games.
method Transformer-based autoregressive model for predicting mesh vertices and faces.
result PolyGen produces high-quality, usable 3D meshes and competitive conditional performance.
Elastic-InfoGAN learns object identity in class-imbalanced data.
problem Learning disentangled representations in class-imbalanced data.
method Invariance to identity-preserving transformations to learn object identity.
result Effectiveness in disentangling object identity in imbalanced datasets.
Proposes a robust 3D classification method for sparse point clouds.
problem Invariance to rotation, positional shift, scaling, and robustness to point sparsity in point cloud classification.
method Introduces a graph-based feature learning approach with an end-to-end neural network.
result Significantly improves 3D object classification and retrieval tasks with sparse point clouds.
The paper proposes a method to learn 3D object pose manifolds using GANs and elasticae.
problem Learning image manifolds of 3D objects with limited data.
method Geom-SGAN and elasticae for geometry-preserving image interpolation.
result The method outperforms state-of-the-art GANs and VAEs in learning rotation paths.
We propose a data-driven 3D shape design method that can learn a generative model from a corpus of existing designs, and use this model to produce a wide range of new designs. The approach learns an encoding of the samples in the training corpus using an unsupervised variational autoencoder-decoder architecture, withou…
Recently, multiple formulations of vision problems as probabilistic inversions of generative models based on computer graphics have been proposed. However, applications to 3D perception from natural images have focused on low-dimensional latent scenes, due to challenges in both modeling and inference. Accounting for th…
GENESIS generates and samples 3D scenes by capturing object interactions.
problem Lack of models that explicitly capture object interactions in scene generation.
method Object-centric latent variables, spatial GMM, amortized inference, autoregressive prior.
result First object-centric generative model of 3D visual scenes.
Paper presents new dataset for disentanglement learning from physical objects.
problem Transfer of disentanglement models from synthetic to real-world data.
method Developed a dataset of physical objects with controlled variations, used a robotic arm for precise manipulation.
result Disentanglement models perform poorly on real data but selection of models and hyperparameters improves transfer.
We introduce SARR for symmetric object pose estimation, improving CNN performance.
problem Ambiguities in symmetric object orientations hinder deep learning pose estimation.
method Numeric rotation representation using symmetry-derived trigonometric identities.
result SARR enables standard CNNs to achieve state-of-the-art performance.
Generates coherent 3D scenes from monocular videos without supervision.
problem Lack of 3D scene modeling in video generation models.
method Trains a model to generate 3D scenes with moving objects and a background from monocular videos.
result Trained model generates coherent 3D scenes with multiple moving objects and a background.
Graph Neural Networks improve 3D object detection in LiDAR point clouds.
problem Challenges in processing LiDAR data due to its 3D geometry and massive volume.
method Proposes a Graph Neural Network (GNN) based framework for 3D object detection.
result GNNs successfully identify objects in 3D LiDAR point clouds.
3D object detection improved using energy-based models.
problem Accurate 3D object detection in cluttered environments from sparse LiDAR data.
method Designing a differentiable pooling operator for 3D bounding boxes integrated into a state-of-the-art 3D object detector.
result Our approach consistently outperforms the SA-SSD baseline across all 3DOD metrics on the KITTI dataset.
The article explains Rao distances and conformal mappings for 3D objects.
problem Calculating distances and preserving angles in 3D objects.
method Proposed constructions of distances and angle-preserving mappings.
result Application to virtual tourism and line integrals in complex planes.
3D Convolutional Neural Networks (3D-CNN) have been used for object recognition based on the voxelized shape of an object. However, interpreting the decision making process of these 3D-CNNs is still an infeasible task. In this paper, we present a unique 3D-CNN based Gradient-weighted Class Activation Mapping method (3D…
Training 3D object detectors for autonomous driving has been limited to small datasets due to the effort required to generate annotations. Reducing both task complexity and the amount of task switching done by annotators is key to reducing the effort and time required to generate 3D bounding box annotations. This paper…
A number of recent approaches to policy learning in 2D game domains have been successful going directly from raw input images to actions. However when employed in complex 3D environments, they typically suffer from challenges related to partial observability, combinatorial exploration spaces, path planning, and a scarc…
Sparsity inducing regularization is an important part for learning over-complete visual representations. Despite the popularity of ℓ1 regularization, in this paper, we investigate the usage of non-convex regularizations in this problem. Our contribution consists of three parts. First, we propose the leaky capped …
Paper develops a differentiable approach for 3D imaging models using Fourier slice theorem.
problem Uncertainty in 3D structure modeling and pose estimation in scientific imaging.
method Differentiable probabilistic models in Fourier space with backpropagation through projection.
result Validates approach on 3D protein reconstruction and extends to probabilistic models.
New method finds Lie group representations without explicit groups, enabling new neural network architectures.
problem Building neural networks equivariant to arbitrary Lie groups.
method Algorithm to find Lie group representations from Lie algebra structure constants. Self-contained method for constructing Lie group-equivariant neural networks.
result First object-tracking model equivariant to the Poincaré group.
A neural scene representation framework enforcing 3D transformations.
problem Learning 3D scene representations from images without 3D supervision.
method Introducing a loss enforcing equivariance of the scene representation with 3D transformations.
result Real-time neural rendering with comparable results to models requiring minutes for inference.
This paper introduces a new approach for the automated reconstruction - reassembly of fragmented objects having one surface near to plane, on the basis of the 3D representation of their constituent fragments. The whole process starts by 3D scanning of the available fragments. The obtained representations are properly p…
3D adversarial logos can fool object detectors in real-world settings.
problem Creating robust adversarial attacks in 3D rendering views.
method Constructing 3D adversarial logos via texture mapping and differentiable rendering.
result 3D adversarial logos are more versatile and robust than traditional adversarial patches.
New method learns stochastic process representations without exact reconstruction.
problem Learning exact representations of high-dimensional noisy stochastic processes.
method CReSP framework for contrastive learning of stochastic processes.
result Effective for learning representations of various stochastic processes.
Generative model disentangles 3D shapes into independent factors.
problem Learning rich representations of deformable 3D shapes.
method Supervised 3D mesh-convolutional Variational AutoEncoder with latent feature disentanglement.
result Explicit disentanglement of latent factors improves shape generation and downstream tasks.
SPACE models complex scenes by decomposing objects and backgrounds.
problem Scalability and unsupervised object-oriented scene representation learning.
method Generative latent variable model combining spatial-attention and scene-mixture approaches.
result SPACE achieves factorized object representations and decomposes complex scenes.
ObSuRF converts a single image into a 3D model with NeRFs.
problem Creating a 3D model from a single image with object segmentation.
method Unsupervised volume segmentation using Neural Radiance Fields (NeRFs).
result ObSuRF can segment a 3D scene into objects from a single image.
Simplified 3D Dijkgraaf-Witten theory with defects explained geometrically.
problem Constructing 3D Dijkgraaf-Witten theory with defects.
method Symmetric monoidal functor from defect cobordism category to vector spaces, using geometric and homotopy theoretic methods.
result Explicit construction of 3D untwisted Dijkgraaf-Witten theory with defects.
Voxel-FPN detects 3D objects from point clouds using raw LIDAR data.
problem 3D object detection in point cloud data for autonomous driving.
method Bottom-up voxel extraction and top-down feature fusion.
result Voxel-FPN outperforms baselines on the KITTI-3D benchmark.
STRING improves 2D and 3D position encodings for better performance.
problem Efficient and accurate position encoding for 2D and 3D applications.
method STRING extends Rotary Position Encodings with a unifying theoretical framework, maintaining translation invariance and low computational cost.
result STRING shows substantial gains in open-vocabulary object detection and robotics.
Cooperative perception improves 3D object detection in autonomous vehicles.
problem Limited field-of-view and occlusion in single sensor data.
method Early fusion of point clouds from multiple sensors, late fusion of independently detected bounding boxes, and hybrid combination.
result Early fusion approach outperforms late fusion by significantly higher recall (95%) compared to single-point sensing (30%).
There are many forms of feature information present in video data. Principle among them are object identity information which is largely static across multiple video frames, and object pose and style information which continuously transforms from frame to frame. Most existing models confound these two types of represen…
Pix2Shape learns 3D scene representations from single images without supervision.
problem Learning 3D scene information from a single image without supervision.
method Pix2Shape uses an encoder, decoder, and critic network to generate 2.5D surfel-based reconstructions.
result Pix2Shape can generate complex 3D scenes from a single image, scaling with on-screen resolution.
This paper considers object detection and 3D estimation using an FMCW radar. The state-of-the-art deep learning framework is employed instead of using traditional signal processing. In preparing the radar training data, the ground truth of an object orientation in 3D space is provided by conducting image analysis, of w…
This paper presents KeypointNet, an end-to-end geometric reasoning framework to learn an optimal set of category-specific 3D keypoints, along with their detectors. Given a single image, KeypointNet extracts 3D keypoints that are optimized for a downstream task. We demonstrate this framework on 3D pose estimation by pro…
Analyze and predict complex 3D shape deformations using LSTM autoencoders and oriented bounding boxes.
problem Detecting and predicting patterns in sequences of deforming 3D shapes.
method Use LSTM autoencoders to create low-dimensional representations of 3D shapes, incorporating oriented bounding boxes for structural components.
result The method detects patterns in plastic deformation and predicts future states of 3D shapes with improved accuracy.