Open Barbend opened 1 year ago
I am training on a dense type of data, like the image in this github page with the streets, cars and people. Am getting a very high number of undetected instances - are there hyper-parameters that i can use for tuning the model in training ?
I am training on a dense type of data, like the image in this github page with the streets, cars and people. Am getting a very high number of undetected instances - are there hyper-parameters that i can use for tuning the model in training ?