The detection result is inconsistent with the expectation

Accelerated Computing Intelligent Video Analytics Visual AI Agent
1

Please provide the following information when creating a topic:

  • Hardware Platform (GPU model and numbers)
    8 * H20
  • System Memory
    1TB
  • Ubuntu Version
    rhel9.4
  • NVIDIA GPU Driver Version (valid for GPU only)
    gpu operator→25.3.0, driver→570.124.06
  • Issue Type( questions, new requirements, bugs)
    bugs
  • How to reproduce the issue ? (

    正例出现误报
    正例出现误报1580×682 153 KB

)
I used a video of lane changing over a solid line to test illegal parking detection, but an alert was triggered, which was not the expected result。
Here is the prompt: Identify the behavior of vehicles parked on a section with no-parking signs in the video, and provide information about the vehicle’s color, license plate number, brand, and model.
Here is the alerts: The Vehicle parked on a section with no-parking signs

3

This requires model fine-tuning based on the specific scenarios. We haven’t provided a fine-tuning plan for now. You can try to fine-tune the VILA-1.5 model by yourself currently.

4

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

5

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.