Linfeng-Tang / SeAFusion

The code of " Image fusion in the loop of high-level vision tasks: A semantic-aware real-time infrared and visible image fusion network"
MIT License
184 stars 34 forks source link

请问能否只训练融合部分的模型? #41

Closed chenmy10 closed 1 month ago

chenmy10 commented 6 months ago

您好,请问如果只想要进行可见光和红外的融合,不需要进行分割,是否可以只训练融合部分模型?是否会对融合模型的性能产生较大的影响?

Zshuting commented 3 months ago

请问你试过吗?可以吗?

chenmy10 commented 2 months ago

请问你试过吗?可以吗?

可以,是两个部分分开训练的。