Hi everyone,
I am working on a detection model with YoloV5. I have trained my custom weights and convert it to tensorRT.
I can achieve around 25-30FPS on a static video. But whenever I enable USB Cam, my FPS decreases to 15-20.
Also without detection, only usbcamera performs quite well.
I tried various GStreamer pipelines but it did not work.
Any idea?