Difference in metrics when changing the clustering algorithm

I trained the model and got the best weights. Now I want to validate the model.

!tao detectnet_v2 evaluate -e $SPECS_DIR/trafficcamnet_finetune.txt \
                           -m $USER_EXPERIMENT_DIR/best_model.tlt \
                           -k $KEY

I am also attaching the part of specs file
trafficcamnet_finetune.txt (2.1 KB)
I also noticed that the value of the nms_confidence_threshold and dbscan_confidence_threshold parameters do not affect the validation metrics.
Peace for everyone.

How about following the jupyter notebook to run KITTI dataset? Will it the same behavior?

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

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