3D perception serves as a cornerstone in the realm of autonomous driving. Vision-based 3D perception methods, which rely solely on camera inputs to reconstruct a 3D environment, have seen significant advancements due to the proliferation of deep learning techniques. Despite these strides, existing frameworks still encounter performance bottlenecks and often necessitate substantial amounts of LiDAR-annotated data, limiting their practical deployment across diverse autonomous driving platforms at a larger scale.

This dissertation is a multifaceted contribution to the advancement of vision-based 3D perception technologies. In the first segment, the thesis introduces structural enhancements to both monocular and stereo 3D object detection algorithms. By integrating ground-referenced geomet...[ Read more ]

3D perception serves as a cornerstone in the realm of autonomous driving. Vision-based 3D perception methods, which rely solely on camera inputs to reconstruct a 3D environment, have seen significant advancements due to the proliferation of deep learning techniques. Despite these strides, existing frameworks still encounter performance bottlenecks and often necessitate substantial amounts of LiDAR-annotated data, limiting their practical deployment across diverse autonomous driving platforms at a larger scale.

This dissertation is a multifaceted contribution to the advancement of vision-based 3D perception technologies. In the first segment, the thesis introduces structural enhancements to both monocular and stereo 3D object detection algorithms. By integrating ground-referenced geometric priors into monocular detection models, this research achieves unparalleled accuracy in benchmark evaluations for monocular 3D detection. Concurrently, the work refines stereo 3D detection paradigms by incorporating insights and inferential structures gleaned from monocular networks, thereby augmenting the operational efficiency of stereo detection systems.

The second segment is devoted to data-driven strategies and their real-world applications in 3D vision detection. A novel training regimen is introduced that amalgamates datasets annotated with either 2D or 3D labels. This approach not only augments the detection models through the utilization of a substantially expanded dataset but also facilitates economical model deployment in real-world scenarios where only 2D annotations are readily available.

Lastly, the dissertation presents an innovative pipeline tailored for unsupervised depth estimation in autonomous driving contexts. Extensive empirical analyses affirm the robustness and efficacy of this newly proposed pipeline. Collectively, these contributions lay a robust foundation for the widespread adoption of vision-based 3D perception technologies in autonomous driving applications.