David Tian

Results 4 issues of David Tian

**Following is the middle results of the training process.** > Val loss: 0.007887, accuray: 0.000049 [17/1000][10/109] Loss: 0.006624 [17/1000][20/109] Loss: 0.006612 [17/1000][30/109] Loss: 0.006566 [17/1000][40/109] Loss: 0.006557 [17/1000][50/109] Loss: 0.006572...

Is there someone who can **share the code** of how to execute the **multi-scale training** for the **citypersons dataset**? Thanks!

预测5个视频,每个3-5分钟以内,会占用内存到爆

VideoTag

如果我想做1倍超分,实质上是图片质量的提升,那训练数据图像对准备的时候,LR和HR是不是尺寸应该一样? 如果是这样的话,那需要使用一些降质方法获得LR图像,请问有合适的降质方法吗?谢谢!