caiyuanhao1998 / MST

A toolbox for spectral compressive imaging reconstruction including MST (CVPR 2022), CST (ECCV 2022), DAUHST (NeurIPS 2022), BiSCI (NeurIPS 2023), HDNet (CVPR 2022), MST++ (CVPRW 2022), etc.
MIT License
571 stars 71 forks source link

为什么测试结果是包含了truth #42

Closed dongruoyu1 closed 7 months ago

dongruoyu1 commented 7 months ago

image 如图所示,pred是4-D的我理解,代表了我们得到的训练结果一共10张图片,每一张都包含28张高光谱图片,那么truth的意义是什么呢?

caiyuanhao1998 commented 7 months ago

是真值,计算PSNR和SSIM用的

dongruoyu1 commented 7 months ago

谢谢您,那么也就是说用不同模型测试得到的pred是不一样的,但是truth都是同一批高光谱图片?

bryanbocao commented 7 months ago

但是truth都是同一批高光谱图片?

@dongruoyu1 Simulation里面是的,论文写Benchmark用KAIST 10 scenes,测试不同模型pred与Grount truth的visual差距,以PSNR和SSIM的metrics来测量。PSNR和SSIM都是reference-based的metrics,以ground truth为reference.