Large scale 3D scene reconstruction is important for applications such as virtual reality and simulation. Existing neural rendering approaches (e.g., NeRF, 3DGS) have achieved realistic reconstructions on large scenes, but optimize per scene, which is expensive and slow, and exhibit noticeable artifacts under large view changes due to overfitting. Generalizable approaches or large reconstruction models are fast, but primarily work for small scenes/objects and often produce lower quality rendering results. In this work, we introduce G3R, a generalizable reconstruction approach that can efficiently predict high-quality 3D scene representations for large scenes. We propose to learn a reconstruction network that takes the gradient feedback signals from differentiable rendering to iteratively update a 3D scene representation, combining the benefits of high photorealism from per-scene optimization with data-driven priors from fast feed-forward prediction methods. Experiments on urban-driving and drone datasets show that G3R generalizes across diverse large scenes and accelerates the reconstruction process by at least 10x while achieving comparable or better realism compared to 3DGS, and also being more robust to large view changes.

Self-driving vehicles (SDVs) require accurate calibration of LiDARs and cameras to fuse sensor data accurately for autonomy. Traditional calibration methods typically leverage fiducials captured in a controlled and structured scene and compute correspondences to optimize over. These approaches are costly and require substantial infrastructure and operations, making it challenging to scale for vehicle fleets. In this work, we propose UniCal, a unified framework for effortlessly calibrating SDVs equipped with multiple LiDARs and cameras. Our approach is built upon a differentiable scene representation capable of rendering multi-view geometrically and photometrically consistent sensor observations. We jointly learn the sensor calibration and the underlying scene representation through differentiable volume rendering, utilizing outdoor sensor data without the need for specific calibration fiducials. This "drive-and-calibrate" approach significantly reduces costs and operational overhead compared to existing calibration systems, enabling efficient calibration for large SDV fleets at scale. To ensure geometric consistency across observations from different sensors, we introduce a novel surface alignment loss that combines feature-based registration with neural rendering. Comprehensive evaluations on multiple datasets demonstrate that UniCal outperforms or matches the accuracy of existing calibration approaches while being more efficient, demonstrating the value of UniCal for scalable calibration.

Large-scale data is crucial for learning realistic and capable driving policies. However, it can be impractical to rely on scaling datasets with real data alone. The majority of driving data is uninteresting, and deliberately collecting new long-tail scenarios is expensive and unsafe. We propose asymmetric self-play to scale beyond real data with additional challenging, solvable, and realistic synthetic scenarios. Our approach pairs a teacher that learns to generate scenarios it can solve but the student cannot, with a student that learns to solve them. When applied to traffic simulation, we learn realistic policies with significantly fewer collisions in both nominal and long-tail scenarios. Our policies further zero-shot transfer to generate training data for end-to-end autonomy, significantly outperforming state-of-the-art adversarial approaches, or using real data alone. For more information, visit https://waabi.ai/selfplay .

The tasks of object detection and trajectory forecasting play a crucial role in understanding the scene for autonomous driving. These tasks are typically executed in a cascading manner, making them prone to compounding errors. Furthermore, there is usually a very thin interface between the two tasks, creating a lossy information bottleneck. To address these challenges, our approach formulates the union of the two tasks as a trajectory refinement problem, where the first pose is the detection (current time), and the subsequent poses are the waypoints of the multiple forecasts (future time). To tackle this unified task, we design a refinement transformer that infers the presence, pose, and multi-modal future behaviors of objects directly from LiDAR point clouds and high-definition maps. We call this model DeTra, short for object Detection and Trajectory forecasting. In our experiments, we observe that \ourmodel{} outperforms the state-of-the-art on Argoverse 2 Sensor and Waymo Open Dataset by a large margin, across a broad range of metrics. Last but not least, we perform extensive ablation studies that show the value of refinement for this task, that every proposed component contributes positively to its performance, and that key design choices were made.

Perceiving the world and forecasting its future state is a critical task for self-driving. Supervised approaches leverage annotated object labels to learn a model of the world -- traditionally with object detections and trajectory predictions, or temporal bird's-eye-view (BEV) occupancy fields. However, these annotations are expensive and typically limited to a set of predefined categories that do not cover everything we might encounter on the road. Instead, we learn to perceive and forecast a continuous 4D (spatio-temporal) occupancy field with self-supervision from LiDAR data. This unsupervised world model can be easily and effectively transferred to downstream tasks. We tackle point cloud forecasting by adding a lightweight learned renderer and achieve state-of-the-art performance in Argoverse 2, nuScenes, and KITTI. To further showcase its transferability, we fine-tune our model for BEV semantic occupancy forecasting and show that it outperforms the fully supervised state-of-the-art, especially when labeled data is scarce. Finally, when compared to prior state-of-the-art on spatio-temporal geometric occupancy prediction, our 4D world model achieves a much higher recall of objects from classes relevant to self-driving.

We are interested in understanding whether retrieval-based localization approaches are good enough in the context of self-driving vehicles. Towards this goal, we introduce Pit30M, a new image and LiDAR dataset with over 30 million frames, which is 10 to 100 times larger than those used in previous work. Pit30M is captured under diverse conditions (i.e., season, weather, time of the day, traffic), and provides accurate localization ground truth. We also automatically annotate our dataset with historical weather and astronomical data, as well as with image and LiDAR semantic segmentation as a proxy measure for occlusion. We benchmark multiple existing methods for image and LiDAR retrieval and, in the process, introduce a simple, yet effective convolutional network-based LiDAR retrieval method that is competitive with the state of the art. Our work provides, for the first time, a benchmark for sub-metre retrieval-based localization at city scale. The dataset, its Python SDK, as well as more information about the sensors, calibration, and metadata, are available on the project website: https://pit30m.github.io/

A self-driving vehicle must understand its environment to determine the appropriate action. Traditional autonomy systems rely on object detection to find the agents in the scene. However, object detection assumes a discrete set of objects and loses information about uncertainty, so any errors compound when predicting the future behavior of those agents. Alternatively, dense occupancy grid maps have been utilized to understand free-space. However, predicting a grid for the entire scene is wasteful since only certain spatio-temporal regions are reachable and relevant to the self-driving vehicle. We present a unified, interpretable, and efficient autonomy framework that moves away from cascading modules that first perceive, then predict, and finally plan. Instead, we shift the paradigm to have the planner query occupancy at relevant spatio-temporal points, restricting the computation to those regions of interest. Exploiting this representation, we evaluate candidate trajectories around key factors such as collision avoidance, comfort, and progress for safety and interpretability. Our approach achieves better highway driving quality than the state-of-the-art in high-fidelity closed-loop simulations.

Learning world models can teach an agent how the world works in an unsupervised manner. Even though it can be viewed as a special case of sequence modeling, progress for scaling world models on robotic applications such as autonomous driving has been somewhat less rapid than scaling language models with Generative Pre-trained Transformers (GPT). We identify two reasons as major bottlenecks: dealing with complex and unstructured observation space, and having a scalable generative model. Consequently, we propose a novel world modeling approach that first tokenizes sensor observations with VQVAE, then predicts the future via discrete diffusion. When applied to learning world models on point cloud observations, our model reduces prior SOTA Chamfer distance by more than 65% for 1s prediction, and more than 50% for 3s prediction, across NuScenes, KITTI Odometry, and Argoverse2 datasets. Our results demonstrate that discrete diffusion on tokenized agent experience can unlock the power of GPT-like unsupervised learning for robotic agents. Figure 1: Our unsupervised world model can produce accurate near-term 1s predictions and diverse multi-future 3s predictions directly on the level of point cloud observations. World models explicitly represent the knowledge of an autonomous agent about its environment. They are defined as a generative model that predicts the next observation in an environment given past observations and the current action. Such a generative model can learn from any unlabeled agent experience, and can be used for both learning and planning in the model-based reinforcement learning framework (Sutton, 1991). This approach has excelled in domains such as Atari (Kaiser et al., 2019), robotic manipulation (Nagabandi et al., 2020), and Minecraft (Hafner et al., 2023). Learning world models can be viewed as a special case of sequence modeling on agent experience. While Generative Pre-trained Transformers (GPT) (Brown et al., 2020) have enabled rapid progress Prediction systems in autonomous driving still require supervised learning, either on the level of bounding boxes (Luo et al., 2018), semantic segmentation (Sadat et al., 2020), or instance segmentation (Hu et al., 2021). However, just as GPT learns to understand language via next token prediction, if a world model can predict unlabeled future observations really well, it must have developed a general understanding of the scene including geometry and dynamics.

Different outdoor illumination conditions drastically alter the appearance of urban scenes, and they can harm the performance of image-based robot perception systems if not seen during training. Camera simulation provides a cost-effective solution to create a large dataset of images captured under different lighting conditions. Towards this goal, we propose LightSim, a neural lighting camera simulation system that enables diverse, realistic, and controllable data generation. LightSim automatically builds lighting-aware digital twins at scale from collected raw sensor data and decomposes the scene into dynamic actors and static background with accurate geometry, appearance, and estimated scene lighting. These digital twins enable actor insertion, modification, removal, and rendering from a new viewpoint, all in a lighting-aware manner. LightSim then combines physically-based and learnable deferred rendering to perform realistic relighting of modified scenes, such as altering the sun location and modifying the shadows or changing the sun brightness, producing spatially- and temporally-consistent camera videos. Our experiments show that LightSim generates more realistic relighting results than prior work. Importantly, training perception models on data generated by LightSim can significantly improve their performance.

Perception systems employed in self-driving vehicles (SDVs) aim to understand the scene both spatially and temporally. Recently, 4D panoptic segmentation has emerged as an important task which involves assigning a semantic label to each observation, as well as an instance ID representing each unique object consistently over time, thus combining semantic segmentation, instance segmentation and object tracking into a single, comprehensive task. Potential applications of this task include building semantic maps, auto-labelling object trajectories, and onboard perception. The task is, however, challenging due to the sparsity of the point-cloud observations, and the computational complexity of 4D spatio-temporal reasoning. Traditionally, researchers have tackled the constituent tasks in isolation, i.e., segmenting classes [1, 2, 3, 4], identifying individual objects [5, 6], and tracking them over time [7, 8]. However, combining multiple networks into a single perception system makes it error-prone, potentially slow, and cumbersome to train.