Skip to main content
eScholarship
Open Access Publications from the University of California

Scholarly Works (3 results)

  • Thesis
  • Peer Reviewed
UCLA Electronic Theses and Dissertations (2024)

Self-driving technologies hold the potential to revolutionize the transportation system. However, current autonomous vehicles predominantly rely on single-agent perception, which is constrained by a limited viewpoint and often suffers from occlusions, sparse sensor measurements, and potential sensor failures. These limitations pose serious safety threats to modern autonomous driving systems, preventing deploying the system at large scale. To overcome these challenges, this thesis advances the field of cooperative perception, enabling connected agents to collaborate and share sensory information, thereby enhancing the robustness and scalability of connected automation.

The first part of this thesis introduces OPV2V, the first large-scale simulation dataset for Vehicle-to-Vehicle cooperative perception. Building on this foundation, it presents a hetero-modal cooperative perception framework, HM-ViT, which employs a novel heterogeneous 3D graph transformer to jointly model inter-agent and intra-agent interactions. This approach enables effective collaboration among agents with distinct sensor modalities, significantly improving the scalability of cooperative perception.

The second part delves into enhancing the robustness of cooperative perception. It introduces V2X-ViT, a unified vision transformer designed to address common V2X challenges such as heterogeneity, localization errors, and time delays. Additionally, it presents V2XP-ASG, an adversarial scene generation framework that employs adversarial attacks to identify challenging scenarios for cooperative perception. These scenarios are subsequently used to fine-tune the model, further improving its robustness.

The final part transitions to the real-world deployment and development of cooperative perception. It introduces V2X-Real, the first large-scale real-world dataset for Vehicle-to-Everything cooperative perception, featuring full sensor modalities and various V2X collaboration modes. Furthermore, it presents V2X-ReaLO, the first online cooperative perception framework for deploying cooperative perception in real-world systems. This framework is benchmarked to provide insights for future research and development in this field.

    • Article
    • Peer Reviewed
    UC San Diego Previously Published Works (2023)
    Statistical learning techniques and increased computational power have facilitated the development of self-driving car technology. However, a limiting factor has been the high expense of scaling and maintaining high-definition (HD) maps. These maps are a crucial backbone for many approaches to self-driving technology. In response to this challenge, we present an approach that fuses pre-built point cloud map data with images to automatically and accurately identify static landmarks such as roads, sidewalks, and crosswalks. Our pipeline utilizes semantic segmentation of 2D images, associates semantic labels with points in point cloud maps to pinpoint locations in the physical world, and employs a confusion matrix formulation to generate a probabilistic bird's-eye view semantic map from semantic point clouds. The approach has been tested in an urban area with different segmentation networks to generate a semantic map with road features. The resulting map provides a rich context of the environment that is valuable for downstream tasks such as trajectory generation and intent prediction. Moreover, it has the potential to be extended to the automatic generation of HD maps for semantic features. The entire software pipeline is implemented in the robot operating system (ROS), a widely used robotics framework, and made available.
      Cover page: Probabilistic Semantic Mapping for Autonomous Driving in Urban EnvironmentsCreative Commons 'BY' version 4.0 license
      • Article
      • Peer Reviewed
      UC San Diego Previously Published Works (2023)
      This work explores methodologies for dynamic trajectory generation for urban driving environments by utilizing coarse global plan representations. In contrast to state-of-the-art architectures for autonomous driving that often leverage lane-level high-definition (HD) maps, we focus on minimizing required map priors that are needed to navigate in dynamic environments that may change over time. To incorporate high-level instructions (i.e., turn right vs. turn left at intersections), we compare various representations provided by lightweight and open-source OpenStreetMaps (OSM) and formulate a conditional generative model strategy to explicitly capture the multimodal characteristics of urban driving. To evaluate the performance of the models introduced, a data collection phase is performed using multiple full-scale vehicles with ground truth labels. Our results show potential use cases in dynamic urban driving scenarios with real-time constraints. The dataset is released publicly as part of this work in combination with code and benchmarks.
        Cover page: Conditional Generative Models for Dynamic Trajectory Generation and Urban DrivingCreative Commons 'BY' version 4.0 license
        Top