Closed proggheli closed 6 months ago
Hi, please refer to issue #1.
How do I regenerated LargeKernel result as you mentioned here, does it require me to retrain it? If that is the case I do not have enough GPU. https://github.com/lixiaoyu2000/Poly-MOT/issues/1#issuecomment-1727018893
Unfortunately, we don't have the bandwidth here to reproduce it, you'll need to generate detections via weight files. You can also ask the official repo to get the .json file.
Okay thank you!
Hi again!
I want to test the validation dataset on various object detection frameworks. So far, I've only managed to do it with CenterPoint because it's the only one I could find with pre-trained detection models in JSON format. The other frameworks seem to have models stored as .pth files. How did you run the validation dataset with LargeKernel3D-L or other detection frameworks?