The recent development of connected and automated vehicle (CAV) technologies has spurred investigations to optimize dense urban traffic, maximizing vehicle speed and throughput. This paper explores advisory autonomy, in which real-time driving advisories are issued to human drivers, thus achieving near-term performance of automated vehicles. Due to the complexity of traffic systems, recent studies of coordinating CAVs have leveraged deep reinforcement learning (RL). Coarse-grained advisory is formalized as zero-order holds, and we consider a range of hold durations from 0.1 to 40 seconds. However, despite the similarity of the higher-frequency tasks for CAVs, a direct application of deep RL fails to generalize to advisory autonomy tasks. To overcome this, we employ zero-shot transfer, training policies on a set of source tasks-specific traffic scenarios with designated hold durations-and then evaluating the efficacy of these policies on different target tasks. We introduce Temporal Transfer Learning (TTL) algorithms to select source tasks for zero-shot transfer, systematically leveraging the temporal structure to solve the full range of tasks. TTL selects the most suitable source tasks to maximize the performance of the range of tasks. We validate our algorithms on diverse mixed-traffic scenarios, demonstrating that TTL more reliably solves the tasks than baselines. This paper underscores the potential of coarse-grained advisory autonomy with TTL in traffic flow optimization.

Transparent object manipulation remains a significant challenge in robotics due to the difficulty of acquiring accurate and dense depth measurements. Conventional depth sensors often fail with transparent objects, resulting in incomplete or erroneous depth data. Existing depth completion methods struggle with interframe consistency and incorrectly model transparent objects as Lambertian surfaces, leading to poor depth reconstruction. To address these challenges, we propose TranSplat, a surface embedding-guided 3D Gaussian Splatting method tailored for transparent objects. TranSplat uses a latent diffusion model to generate surface embeddings that provide consistent and continuous representations, making it robust to changes in viewpoint and lighting. By integrating these surface embeddings with input RGB images, TranSplat effectively captures the complexities of transparent surfaces, enhancing the splatting of 3D Gaussians and improving depth completion. Evaluations on synthetic and real-world transparent object benchmarks, as well as robot grasping tasks, show that TranSplat achieves accurate and dense depth completion, demonstrating its effectiveness in practical applications. We open-source synthetic dataset and model: https://github.com/jeongyun0609/TranSplat

There has been considerable interest in recent years regarding how a small fraction of autonomous vehicles (AVs) can mitigate traffic congestion. However, the reality of vehicle-based congestion mitigation remains elusive, due to challenges of cost, technology maturity, and regulation. As a result, recent works have investigated the necessity of autonomy by exploring driving advisory systems. Such early works have made simplifying assumptions such as perfect driver compliance. This work relaxes this assumption, focusing on compliance errors caused by physical limitations of human drivers, in particular, response delay and speed deviation. These compliance errors introduce significant unpredictability into traffic systems, complicating the design of real-time driving advisories aimed at stabilizing traffic flow. Our analysis reveals that performance degradation increases sharply under compliance errors, highlighting the associated difficulties. To address this challenge, we develop a reinforcement learning (RL) framework based on an action-persistent Markov decision process (MDP) combined with domain randomization, designed for robust coarse-grained driving policies. This approach allows driving policies to effectively manage the cumulative impacts of compliance errors by generating various scenarios and corresponding traffic conditions during training. We show that in comparison to prior RL-based work which did not consider compliance errors, our policies achieve up to 2.2 times improvement in average speed over non-robust training. In addition, analytical results validate the experiment results, highlighting the benefits of the proposed framework. Overall, this paper advocates the necessity of incorporating human driver compliance errors in the development of RL-based advisory systems, achieving more effective and resilient traffic management solutions.

The massive amount of data related to spatiotemporal mobility offers new opportunities to understand human mobility with applications in various sectors, including transportation, logistics, and safety. However, the increase in the volume and in the dimension of mobility data makes it challenging to retrieve important information and critical features of spatiotemporal mobility. This paper develops a method to estimate probabilistic occurrences of travel demands considering interactions between origin, destination, and departure time. First, we reveal the important features in the complex structure of mobility data and identify mobility patterns. Then, we derive a data-driven model, accounting for mobility patterns, to estimate and predict travel demands. We quantify the accuracy of the proposed method for a case study using both New York city yellow taxi trip data and for-hire vehicles trip data over the entire city. Results show the accuracy of the proposed method compared to existing approaches.

The massive amount of data related to spatiotem-poral mobility offers new opportunities to understand human behaviors. However, with the increase of volume and complexity of mobility data, it has become challenging to retrieve important information and critical features of spatiotemporal mobility. In particular, predicting large-scale travel demands is challenging and requires a high computational load. This paper introduces a data-driven approach for estimating high-dimensional travel demands. We propose a method to identify mobility patterns using a probabilistic tensor decomposition approach for interpreting the complexity and uncertainty of mobility data. Expectation-maximization (EM) algorithm is applied for inferring mobility patterns. A case study is presented, where the proposed model is applied to New York city taxi data. The results show the model performance according to the number of origin and destination patterns and the number of trip data used. The probabilistic modeling results provide a deeper understanding of large-scale mobility data in the spatiotemporal dimension.

This paper introduces a stereo thermal camera dataset (STheReO) with multiple navigation sensors to encourage thermal SLAM researches. A thermal camera measures infrared rays beyond the visible spectrum therefore it could provide a simple yet robust solution to visually degraded environments where existing visual sensor-based SLAM would fail. Existing thermal camera datasets mostly focused on monocular configuration using the thermal camera with RGB cameras in a visually challenging environment. A few stereo thermal rig were examined but in computer vision perspective without supporting sequential images for state estimation algorithms. To encourage the academia for the evolving stereo thermal SLAM, we obtain nine sequences in total across three spatial locations and three different times per location (e.g., morning, day, and night) to capture the variety of thermal characteristics. By using the STheReO dataset, we hope diverse types of researches will be made, including but not limited to odometry, mapping, and SLAM (e.g., thermal-LiDAR mapping or long-term thermal localization). Our datasets are available at https://sites.google.com/view/rpmsthereo/.

Today, people use a computer almost everywhere. At the same time, they still do their work in the old-fashioned way, such as using a pen and paper. A pen is often used in many fields because it is easy to use and familiar. On the other hand, however, it is a quite inconvenient because the information printed on paper is static. If digital features are added to this paper environment, the users can do their work more easily and efficiently. AR (augmented reality) Lamp is a stand-type projector and camera embedded system with the form factor of a desk lamp. Its users can modify the virtually augmented content on top of the paper with seamlessly combined virtual and physical worlds. AR is quite appealing, but it is difficult to popularize due to the lack of interaction. In this paper, the interaction methods that people can use easily and intuitively are focused on. A high-fidelity prototype of the system is presented, and a set of novel interactions is demonstrated. A pilot evaluation of the system is also reported to explore its usage possibility.