There are multiple options to think and explore for an improved binary classifier:
- Have a non-moth class in the fine-grained classifier
- Understand the classes where the current model is getting wrong and include those in the training dataset
- Fine-tune the AMI-GBIF trained model with AMI-Traps binary classification data
Make use of https://3lc.ai/
There are multiple options to think and explore for an improved binary classifier:
Make use of https://3lc.ai/