[ICCV2019] Joint Monocular 3D Vehicle Detection and Tracking

แชร์
ฝัง
  • เผยแพร่เมื่อ 13 ต.ค. 2024
  • Website: eborboihuc.git...
    Code: github.com/ucb...

ความคิดเห็น • 5

  • @kwea123
    @kwea123 5 ปีที่แล้ว +3

    4:16 love it when the cars keep the same ids after getting occluded and reappear. In KITTI's evaluation it is not counted as an id switch if it changes id after getting occluded; many papers didn't treat or even recognize this problem. Great job.

  • @samirelzein1095
    @samirelzein1095 2 ปีที่แล้ว

    Great work, it would also be useful to have a live video while installing and coding/using the package with an input video, so that we see better the big picture and deploy your code quick. Congrats again.

  • @DanHaiduc
    @DanHaiduc 5 ปีที่แล้ว +1

    3:54 missed the bus! heheh
    Still, amazing work!

  • @berkantay790
    @berkantay790 3 ปีที่แล้ว

    How did you gather the distance and the 3d box annotation data from the GTAV? Did you label it manually?

    • @williamclifford657
      @williamclifford657 3 ปีที่แล้ว +1

      They answer this in their paper:
      "Our simulation is based on Grand Theft Auto V, a modern
      game that simulates a functioning city and its surroundings
      in a photo-realistic three-dimensional world. To associate
      object instances across frames, we utilize in-game API to
      capture global instance id and corresponding 3D annotations
      directly."