sedna
sedna copied to clipboard
How do sedna conform if is hard sample in incremental_learning?
Sedna provide 3 general algorithms for hard sample mining in the common CV scenario.
- more detail : ('ThresholdFilter', 'CrossEntropyFilter', 'IBTFilter') -> https://support.huaweicloud.com/devg-hilens/hilens_05_0068.html
You can define new algorithms to support more scenarios as you need by follow here.