“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告

A morphable model for the synthesis of 3D faces

We present Block-NeRF, a variant of Neural Radiance Fields that can represent large-scale environments. Specifically, we demonstrate that when scaling NeRF to render city-scale scenes spanning multiple blocks, it is vital to de-compose the scene into individually trained NeRFs. This decomposition decouples rendering time from scene size, enables rendering to scale to arbitrarily large environments, and allows per-block updates of the environment. We adopt several architectural changes to make NeRF robust to data captured over months under different environmental conditions. We add appearance embeddings, learned pose refinement, and controllable exposure to each individual NeRF, and introduce a procedure for aligning appearance between adjacent NeRFs so that they can be seamlessly combined. We build a grid of Block-NeRFs from 2.8 million images to create the largest neural scene representation to date, capable of rendering an entire neighborhood of San Francisco.

Block-NeRF: Scalable Large Scene Neural View Synthesis

We present a new framework for interior scene synthesis that combines a high-level relation graph representation with spatial prior neural networks. We observe that prior work on scene synthesis is divided into two camps: object-oriented approaches (which reason about the set of objects in a scene and their configurations) and space-oriented approaches (which reason about what objects occupy what regions of space). Our insight is that the object-oriented paradigm excels at high-level planning of how a room should be laid out, while the space-oriented paradigm performs well at instantiating a layout by placing objects in precise spatial configurations. With this in mind, we present PlanIT, a layout-generation framework that divides the problem into two distinct planning and instantiation phases. PlanIT represents the "plan" for a scene via a relation graph, encoding objects as nodes and spatial/semantic relationships between objects as edges. In the planning phase, it uses a deep graph convolutional generative model to synthesize relation graphs. In the instantiation phase, it uses image-based convolutional network modules to guide a search procedure that places objects into the scene in a manner consistent with the graph. By decomposing the problem in this way, PlanIT generates scenes of comparable quality to those generated by prior approaches (as judged by both people and learned classifiers), while also providing the modeling flexibility of the intermediate relationship graph representation. These graphs allow the system to support applications such as scene synthesis from a partial graph provided by a user.

/pdf/planit-planning-and-instantiating-indoor-scenes-with-51wnaenft8.pdf

PlanIT: planning and instantiating indoor scenes with relation graph and spatial prior networks

We introduce PREDATOR, a model for pairwise point-cloud registration with deep attention to the overlap region. Different from previous work, our model is specifically designed to handle (also) point-cloud pairs with low overlap. Its key novelty is an overlap-attention block for early information exchange between the latent encodings of the two point clouds. In this way the subsequent decoding of the latent representations into per-point features is conditioned on the respective other point cloud, and thus can predict which points are not only salient, but also lie in the overlap region between the two point clouds. The ability to focus on points that are relevant for matching greatly improves performance: PREDATOR raises the rate of successful registrations by more than 20% in the low-overlap scenario, and also sets a new state of the art for the 3DMatch benchmark with 89% registration recall.

PREDATOR: Registration of 3D Point Clouds with Low Overlap

Autonomous driving system development is critically dependent on the ability to replay complex and diverse traffic scenarios in simulation. In such scenarios, the ability to accurately simulate the vehicle sensors such as cameras, lidar or radar is hugely helpful. However, current sensor simulators leverage gaming engines such as Unreal or Unity, requiring manual creation of environments, objects, and material properties. Such approaches have limited scalability and fail to produce realistic approximations of camera, lidar, and radar data without significant additional work. In this paper, we present a simple yet effective approach to generate realistic scenario sensor data, based only on a limited amount of lidar and camera data collected by an autonomous vehicle. Our approach uses texture-mapped surfels to efficiently reconstruct the scene from an initial vehicle pass or set of passes, preserving rich information about object 3D geometry and appearance, as well as the scene conditions. We then leverage a SurfelGAN network to reconstruct realistic camera images for novel positions and orientations of the self-driving vehicle and moving objects in the scene. We demonstrate our approach on the Waymo Open Dataset and show that it can synthesize realistic camera data for simulated scenarios. We also create a novel dataset that contains cases in which two self-driving vehicles observe the same scene at the same time. We use this dataset to provide additional evaluation and demonstrate the usefulness of our SurfelGAN model.

/pdf/surfelgan-synthesizing-realistic-sensor-data-for-autonomous-577sojbpwi.pdf

SurfelGAN: Synthesizing Realistic Sensor Data for Autonomous Driving

We present a deep generative scene modeling technique for indoor environments. Our goal is to train a generative model using a feed-forward neural network that maps a prior distribution (e.g., a normal distribution) to the distribution of primary objects in indoor scenes. We introduce a 3D object arrangement representation that models the locations and orientations of objects, based on their size and shape attributes. Moreover, our scene representation is applicable for 3D objects with different multiplicities (repetition counts), selected from a database. We show a principled way to train this model by combining discriminative losses for both a 3D object arrangement representation and a 2D image-based representation. We demonstrate the effectiveness of our scene representation and the network training method on benchmark datasets. We also show the applications of this generative model in scene interpolation and scene completion.

Deep Generative Modeling for Scene Synthesis via Hybrid Representations

This paper advocates the use of implicit surface representation in autoencoder-based self-supervised 3D representation learning. The most popular and accessible 3D representation, i.e., point clouds, involves discrete samples of the underlying continuous 3D surface. This discretization process introduces sampling variations on the 3D shape, making it challenging to develop transferable knowledge of the true 3D geometry. In the standard autoencoding paradigm, the encoder is compelled to encode not only the 3D geometry but also information on the specific discrete sampling of the 3D shape into the latent code. This is because the point cloud reconstructed by the decoder is considered unacceptable unless there is a perfect mapping between the original and the reconstructed point clouds. This paper introduces the Implicit AutoEncoder (IAE), a simple yet effective method that addresses the sampling variation issue by replacing the commonly-used point-cloud decoder with an implicit decoder. The implicit decoder reconstructs a continuous representation of the 3D shape, independent of the imperfections in the discrete samples. Extensive experiments demonstrate that the proposed IAE achieves state-of-the-art performance across various self-supervised learning benchmarks.

Implicit Autoencoder for Point Cloud Self-supervised Representation Learning

In this paper, we introduce a novel RGB-D based relative pose estimation approach that is suitable for small-overlapping or non-overlapping scans and can output multiple relative poses. Our method performs scene completion and matches the completed scans. However, instead of using a fixed representation for completion, the key idea is to utilize hybrid representations that combine 360-image, 2D image-based layout, and planar patches. This approach offers adaptively feature representations for relative pose estimation. Besides, we introduce a global-2-local matching procedure, which utilizes initial relative poses obtained during the global phase to detect and then integrate geometric relations for pose refinement. Experimental results justify the potential of this approach across a wide range of benchmark datasets. For example, on ScanNet, the rotation translation errors of the top-1/top-5 predictions of our approach are $28.6^{\circ}/0.90m$ and $16.8^{\circ}/0.76m$, respectively. Our approach also considerably boosts the performance of multi-scan reconstruction in few-view reconstruction settings.

/pdf/extreme-relative-pose-network-under-hybrid-representations-56ry3p9auj.pdf

Extreme Relative Pose Network Under Hybrid Representations

Estimating the relative rigid pose between two RGB-D scans of the same underlying environment is a fundamental problem in computer vision, robotics, and computer graphics. Most existing approaches allow only limited maximum relative pose changes since they require considerable overlap between the input scans. We introduce a novel approach that extends the scope to extreme relative poses, with little or even no overlap between the input scans. The key idea is to infer more complete scene information about the underlying environment and match on the completed scans. In particular, instead of only performing scene completion from each individual scan, our approach alternates between relative pose estimation and scene completion. This allows us to perform scene completion by utilizing information from both input scans at late iterations, resulting in better results for both scene completion and relative pose estimation. Experimental results on benchmark datasets show that our approach leads to considerable improvements over state-of-the-art approaches for relative pose estimation. In particular, our approach provides encouraging relative pose estimates even between non-overlapping scans.

/pdf/extreme-relative-pose-estimation-for-rgb-d-scans-via-scene-582x1nz2u9.pdf

Zhenpei Yang

Papers

SurfelGAN: Synthesizing Realistic Sensor Data for Autonomous Driving

Deep Generative Modeling for Scene Synthesis via Hybrid Representations

Implicit Autoencoder for Point Cloud Self-supervised Representation Learning

Extreme Relative Pose Network Under Hybrid Representations

Extreme Relative Pose Estimation for RGB-D Scans via Scene Completion