Open alvinhuai opened 1 year ago
I used this algorithm to train a model containing 200 detection targets. Some of these detection targets are relatively similar, but the similarity in actual recognition is very low. In many cases, A is recognized as B. How to solve this?
I used this algorithm to train a model containing 200 detection targets. Some of these detection targets are relatively similar, but the similarity in actual recognition is very low. In many cases, A is recognized as B. How to solve this?