Witnessing the impressive achievements of pre-training techniques on large-scale data in the field of computer vision and natural language processing, we wonder whether this idea could be adapted in a grab-and-go spirit, and mitigate the sample inefficiency problem for visuomotor driving. Given the highly dynamic and variant nature of the input, the visuomotor driving task inherently lacks view and translation invariance, and the visual input contains massive irrelevant information for decision making, resulting in predominant pre-training approaches from general vision less suitable for the autonomous driving task. To this end, we propose PPGeo (Policy Pre-training via Geometric modeling), an intuitive and straightforward fully self-supervised framework curated for the policy pretraining in visuomotor driving. We aim at learning policy representations as a powerful abstraction by modeling 3D geometric scenes on large-scale unlabeled and uncalibrated YouTube driving videos. The proposed PPGeo is performed in two stages to support effective self-supervised training. In the first stage, the geometric modeling framework generates pose and depth predictions simultaneously, with two consecutive frames as input. In the second stage, the visual encoder learns driving policy representation by predicting the future ego-motion and optimizing with the photometric error based on current visual observation only. As such, the pre-trained visual encoder is equipped with rich driving policy related representations and thereby competent for multiple visuomotor driving tasks. Extensive experiments covering a wide span of challenging scenarios have demonstrated the superiority of our proposed approach, where improvements range from 2% to even over 100% with very limited data. Code and models will be available at https://github.com/OpenDriveLab/PPGeo.

Graph neural networks (GNNs), as the de-facto model class for representation learning on graphs, are built upon the multi-layer perceptrons (MLP) architecture with additional message passing layers to allow features to flow across nodes. While conventional wisdom largely attributes the success of GNNs to their advanced expressivity for learning desired functions on nodes' ego-graphs, we conjecture that this is \emph{not} the main cause of GNNs' superiority in node prediction tasks. This paper pinpoints the major source of GNNs' performance gain to their intrinsic generalization capabilities, by introducing an intermediate model class dubbed as P(ropagational)MLP, which is identical to standard MLP in training, and then adopt GNN's architecture in testing. Intriguingly, we observe that PMLPs consistently perform on par with (or even exceed) their GNN counterparts across ten benchmarks and different experimental settings, despite the fact that PMLPs share the same (trained) weights with poorly-performed MLP. This critical finding opens a door to a brand new perspective for understanding the power of GNNs, and allow bridging GNNs and MLPs for dissecting their generalization behaviors. As an initial step to analyze PMLP, we show its essential difference with MLP at infinite-width limit lies in the NTK feature map in the post-training stage. Moreover, though MLP and PMLP cannot extrapolate non-linear functions for extreme OOD data, PMLP has more freedom to generalize near the training support.

Contrastive learning methods based on InfoNCE loss are popular in node representation learning tasks on graph-structured data. However, its reliance on data augmentation and its quadratic computational complexity might lead to inconsistency and inefficiency problems. To mitigate these limitations, in this paper, we introduce a simple yet effective contrastive model named Localized Graph Contrastive Learning (Local-GCL in short). Local-GCL consists of two key designs: 1) We fabricate the positive examples for each node directly using its first-order neighbors, which frees our method from the reliance on carefully-designed graph augmentations; 2) To improve the efficiency of contrastive learning on graphs, we devise a kernelized contrastive loss, which could be approximately computed in linear time and space complexity with respect to the graph size. We provide theoretical analysis to justify the effectiveness and rationality of the proposed methods. Experiments on various datasets with different scales and properties demonstrate that in spite of its simplicity, Local-GCL achieves quite competitive performance in self-supervised node representation learning tasks on graphs with various scales and properties.

Deep neural networks still struggle on long-tailed image datasets, and one of the reasons is that the imbalance of training data across categories leads to the imbalance of trained model parameters. Motivated by the empirical findings that trained classifiers yield larger weight norms in head classes, we propose to reformulate the recognition probabilities through included angles without re-balancing the classifier weights. Specifically, we calculate the angles between the data feature and the class-wise classifier weights to obtain angle-based prediction results. Inspired by the performance improvement of the predictive form reformulation and the outstanding performance of the widely used two-stage learning framework, we explore the different properties of this angular prediction and propose novel modules to improve the performance of different components in the framework. Our method is able to obtain the best performance among peer methods without pretraining on CIFAR10/100-LT and ImageNet-LT. Source code will be made publicly available.

Most existing person re-identification methods compute the matching relations between person images across camera views based on the ranking of the pairwise similarities. This matching strategy with the lack of the global viewpoint and the context's consideration inevitably leads to ambiguous matching results and sub-optimal performance. Based on a natural assumption that images belonging to the same person identity should not match with images belonging to multiple different person identities across views, called the unicity of person matching on the identity level, we propose an end-to-end person unicity matching architecture for learning and refining the person matching relations. First, we adopt the image samples' contextual information in feature space to generate the initial soft matching results by using graph neural networks. Secondly, we utilize the samples' global context relationship to refine the soft matching results and reach the matching unicity through bipartite graph matching. Given full consideration to real-world person re-identification applications, we achieve the unicity matching in both one-shot and multi-shot settings of person re-identification and further develop a fast version of the unicity matching without losing the performance. The proposed method is evaluated on five public benchmarks, including four multi-shot datasets MSMT17, DukeMTMC, Market1501, CUHK03, and a one-shot dataset VIPeR. Experimental results show the superiority of the proposed method on performance and efficiency.

现有检测方法通常使用参数化边界框（Bbox）进行建模和检测（水平）对象，并将其他旋转角参数用于旋转对象。我们认为，这种机制在建立有效的旋转检测回归损失方面具有根本的局限性，尤其是对于高精度检测而言，高精度检测（例如0.75）。取而代之的是，我们建议将旋转的对象建模为高斯分布。一个直接的优势是，我们关于两个高斯人之间距离的新回归损失，例如kullback-leibler Divergence（KLD）可以很好地对齐实际检测性能度量标准，这在现有方法中无法很好地解决。此外，两个瓶颈，即边界不连续性和正方形的问题也消失了。我们还提出了一种有效的基于高斯度量的标签分配策略，以进一步提高性能。有趣的是，通过在基于高斯的KLD损失下分析Bbox参数的梯度，我们表明这些参数通过可解释的物理意义进行了动态更新，这有助于解释我们方法的有效性，尤其是对于高精度检测。我们使用量身定制的算法设计将方法从2-D扩展到3-D，以处理标题估计，并在十二个公共数据集（2-D/3-D，空中/文本/脸部图像）上进行了各种基本检测器的实验结果。展示其优越性。

在鸟眼中学习强大的表现（BEV），以进行感知任务，这是趋势和吸引行业和学术界的广泛关注。大多数自动驾驶算法的常规方法在正面或透视视图中执行检测，细分，跟踪等。随着传感器配置变得越来越复杂，从不同的传感器中集成了多源信息，并在统一视图中代表功能至关重要。 BEV感知继承了几个优势，因为代表BEV中的周围场景是直观和融合友好的。对于BEV中的代表对象，对于随后的模块，如计划和/或控制是最可取的。 BEV感知的核心问题在于（a）如何通过从透视视图到BEV来通过视图转换来重建丢失的3D信息； （b）如何在BEV网格中获取地面真理注释； （c）如何制定管道以合并来自不同来源和视图的特征； （d）如何适应和概括算法作为传感器配置在不同情况下各不相同。在这项调查中，我们回顾了有关BEV感知的最新工作，并对不同解决方案进行了深入的分析。此外，还描述了该行业的BEV方法的几种系统设计。此外，我们推出了一套完整的实用指南，以提高BEV感知任务的性能，包括相机，激光雷达和融合输入。最后，我们指出了该领域的未来研究指示。我们希望该报告能阐明社区，并鼓励对BEV感知的更多研究。我们保留一个活跃的存储库来收集最新的工作，并在https://github.com/openperceptionx/bevperception-survey-recipe上提供一包技巧的工具箱。

许多现有的自动驾驶范式涉及多个任务的多个阶段离散管道。为了更好地预测控制信号并增强用户安全性，希望从联合时空特征学习中受益的端到端方法是可取的。尽管基于激光雷达的输入或隐式设计有一些开创性的作品，但在本文中，我们在可解释的基于视觉的设置中提出了问题。特别是，我们提出了一种空间性特征学习方案，以同时同时进行感知，预测和计划任务的一组更具代表性的特征，称为ST-P3。具体而言，提出了一种以自我为中心的积累技术来保留3D空间中的几何信息，然后才能感知鸟类视图转化。设计了双重途径建模，以考虑将来的预测，以将过去的运动变化考虑到过去。引入了基于时间的精炼单元，以弥补识别基于视觉的计划的元素。据我们所知，我们是第一个系统地研究基于端视力的自主驾驶系统的每个部分。我们在开环Nuscenes数据集和闭环CARLA模拟上对以前的最先进的方法进行基准测试。结果显示了我们方法的有效性。源代码，模型和协议详细信息可在https://github.com/openperceptionx/st-p3上公开获得。

配备了广泛的传感器，主要的自主驾驶解决方案正变得越来越面向安全系统设计。尽管这些传感器已经奠定了坚实的基础，但最新的大多数生产解决方案仍然属于L2阶段。其中，Comma.ai出现在我们的视线中，声称一个售价999美元的售后设备装有单个相机和内部的木板具有处理L2场景的能力。该项目与Comma.ai发布的整个系统的开源软件一起名为OpenPilot。可能吗？如果是这样，它如何成为可能？考虑到好奇心，我们深入研究了OpenPilot，并得出结论，其成功的关键是端到端系统设计，而不是传统的模块化框架。该模型被简要介绍为SuperCombo，它可以从单眼输入中预测自我车辆的未来轨迹和其他道路语义。不幸的是，无法公开提供所有这些工作的培训过程和大量数据。为了进行深入的调查，我们尝试重新实现培训细节并测试公共基准测试的管道。这项工作中提出的重构网络称为“ op-Deepdive”。为了将我们的版本与原始SuperCombo进行公平的比较，我们引入了双模型部署方案，以测试现实世界中的驾驶性能。 Nuscenes，Comma2K19，Carla和内部现实场景的实验结果证明了低成本设备确实可以实现大多数L2功能，并且与原始的SuperCombo模型相当。在本报告中，我们想分享我们的最新发现，并阐明了从工业产品级别方面进行端到端自动驾驶的新观点，并有可能激发社区继续提高绩效。我们的代码，基准在https://github.com/openperceptionx/openpilot-deepdive上。

当前的端到端自动驾驶方法要么基于计划的轨迹运行控制器，要么直接执行控制预测，这已经跨越了两条单独研究的研究线。本文看到了它们彼此的潜在相互利益，主动探讨了这两个发展良好的世界的结合。具体而言，我们的集成方法分别有两个用于轨迹计划和直接控制的分支。轨迹分支可以预测未来的轨迹，而控制分支则涉及一种新颖的多步预测方案，以便可以将当前动作与未来状态之间的关系进行推理。连接了两个分支，因此控制分支在每个时间步骤中从轨迹分支接收相应的指导。然后将来自两个分支的输出融合以实现互补的优势。我们的结果在闭环城市驾驶环境中进行了评估，并使用CARLA模拟器具有挑战性的情况。即使有了单眼相机的输入，建议的方法在官方Carla排行榜上排名第一$，超过了其他具有多个传感器或融合机制的复杂候选人。源代码和数据将在https://github.com/openperceptionx/tcp上公开提供。