Urban intersection navigation
VideoPrompt Response Writing SFT
Training a real-time computer version model for autonomous navigation at urban intersections, the model must identify, track and interpret objects to make millisecond decisions.
Specific Data labeling tasks perfumed; 2D bounding boxes, 3D cuboids, Semantic segmentation, Keypoint labeling.
Project size; Thousands of hours of high definition video footage, Paired with LiDAR point clouds.
quality measures; High precision, data scrubbing, Edge case focus and Active learning.
Training a real-time computer version model for autonomous navigation at urban intersections, the model must identify, track and interpret objects to make millisecond decisions. Specific Data labeling tasks perfumed; 2D bounding boxes, 3D cuboids, Semantic segmentation, Keypoint labeling. Project size; Thousands of hours of high definition video footage, Paired with LiDAR point clouds. quality measures; High precision, data scrubbing, Edge case focus and Active learning.
2023 - 2026