Search

Smart traffic control and management become an emerging application for Deep Reinforcement Learning (DRL) to solve traffic congestion problems in urban networks. Different traffic control and management policies can be tested on the traffic simulation. Current DRL-based studies are mainly supported by the microscopic simulation software (e.g., SUMO), while it is not suitable for city-wide control due to the computational burden and gridlock effect. To the best of our knowledge, there is a lack of studies on the large-scale traffic simulator for DRL testbeds, which could further hinder the development of DRL. In view of this, we propose a meso-macro traffic simulator for very large-scale DRL scenarios. The proposed simulator integrates mesoscopic and macroscopic traffic simulation models to improve efficiency and eliminate gridlocks. The mesoscopic link model simulates flow dynamics on roads, and the macroscopic Bathtub model depicts vehicle movement in regions. Moreover, both types of models can be hybridized to accommodate various DRL tasks. This creates portals for mixed transportation applications under different contexts. The result shows that the developed simulator only takes 46 seconds to finish a 24-hour simulation in a very large city with 2.2 million vehicles, which is much faster than SUMO. Additionally, we develop a graphic interface for users to visualize the simulation results in a web explorer. In the future, the developed meso-macro traffic simulator could serve as a new environment for very large-scale DRL problems., Comment: IJCAI 2021 Reinforcement Learning for Intelligent Transportation Systems (RL4ITS) Workshop

Recovering the spatial layout of the cameras and the geometry of the scene from extreme-view images is a longstanding challenge in computer vision. Prevailing 3D reconstruction algorithms often adopt the image matching paradigm and presume that a portion of the scene is co-visible across images, yielding poor performance when there is little overlap among inputs. In contrast, humans can associate visible parts in one image to the corresponding invisible components in another image via prior knowledge of the shapes. Inspired by this fact, we present a novel concept called virtual correspondences (VCs). VCs are a pair of pixels from two images whose camera rays intersect in 3D. Similar to classic correspondences, VCs conform with epipolar geometry; unlike classic correspondences, VCs do not need to be co-visible across views. Therefore VCs can be established and exploited even if images do not overlap. We introduce a method to find virtual correspondences based on humans in the scene. We showcase how VCs can be seamlessly integrated with classic bundle adjustment to recover camera poses across extreme views. Experiments show that our method significantly outperforms state-of-the-art camera pose estimation methods in challenging scenarios and is comparable in the traditional densely captured setup. Our approach also unleashes the potential of multiple downstream tasks such as scene reconstruction from multi-view stereo and novel view synthesis in extreme-view scenarios., CVPR 2022. Project page: https://people.csail.mit.edu/weichium/virtual-correspondence/

We present Neural Mixtures of Planar Experts (NeurMiPs), a novel planar-based scene representation for modeling geometry and appearance. NeurMiPs leverages a collection of local planar experts in 3D space as the scene representation. Each planar expert consists of the parameters of the local rectangular shape representing geometry and a neural radiance field modeling the color and opacity. We render novel views by calculating ray-plane intersections and composite output colors and densities at intersected points to the image. NeurMiPs blends the efficiency of explicit mesh rendering and flexibility of the neural radiance field. Experiments demonstrate superior performance and speed of our proposed method, compared to other 3D representations in novel view synthesis., CVPR 2022. Project page: https://zhihao-lin.github.io/neurmips/

Neural Radiance Fields (NeRF) have recently gained a surge of interest within the computer vision community for its power to synthesize photorealistic novel views of real-world scenes. One limitation of NeRF, however, is its requirement of accurate camera poses to learn the scene representations. In this paper, we propose Bundle-Adjusting Neural Radiance Fields (BARF) for training NeRF from imperfect (or even unknown) camera poses -- the joint problem of learning neural 3D representations and registering camera frames. We establish a theoretical connection to classical image alignment and show that coarse-to-fine registration is also applicable to NeRF. Furthermore, we show that na\"ively applying positional encoding in NeRF has a negative impact on registration with a synthesis-based objective. Experiments on synthetic and real-world data show that BARF can effectively optimize the neural scene representations and resolve large camera pose misalignment at the same time. This enables view synthesis and localization of video sequences from unknown camera poses, opening up new avenues for visual localization systems (e.g. SLAM) and potential applications for dense 3D mapping and reconstruction., Comment: Accepted to ICCV 2021 as oral presentation (project page & code: https://chenhsuanlin.bitbucket.io/bundle-adjusting-NeRF)

The rapid development of cooperative vehicle-infrastructure system (CVIS) improves the communication reliability between vehicles and road environment. These communications enable the accurate vehicle rollover prediction in Human-Vehicle-Road interaction. However, considering the strong non-linear characteristics of Human-Vehicle-Road interaction and the uncertainty of modeling, the traditional deterministic method cannot meet the requirement of accurate prediction of rollover hazard for heavy vehicles. In order to improve the accuracy of vehicles rollover prediction, this paper proposes a developed rollover prediction algorithm based on the multiple observed variables by combining the failure probability in reliability and the empirical model. This approach applies the probability method of uncertainty to the design of dynamic rollover prediction algorithm for heavy vehicles and establishes a classification model of heavy vehicles based on support vector machine (SVM) with multiple observed variables. The failure probability of rollover limit state of heavy vehicles is calculated by Monte Carlo Sampling (MCS), Radial-Based Importance Sampling (RBIS), and Truncated Importance Sampling (TIS), respectively. Then the Fishhook, Double Lane Change tests, and J-turn tests, simulated in TruckSim, are carried out to validate the proposed algorithm. The simulation results show that the rollover prediction algorithm based on failure probability can effectively improve the rollover prediction accuracy for heavy vehicles. Moreover, based on the communication in CVIS, the failure probability can be obtained before entering the specific road. Meanwhile, this approach can reduce the external interference of strong non-linear characteristics of Human-Vehicle-Road interaction and the uncertainty of the modeling to the system, thus improving the prediction accuracy of active safety performance of heavy vehicles significantly.

A new kind of vehicle-mounted battlefield perception system design and implementation scheme is presented, battlefield situation perception and target detection range can be extended by 360°panorama and multi-light fusion imaging technology; system imaging stability and reliability can be improved by multi-axis stabilized platform design; target recognition efficiency and detection and tracking accuracy in battlefield complex conditions can be realized by integrated target detection and tracking algorithm, land based mobile weapon platform battlefield perception ability is improved comprehensively.