北邮-刘畅
北邮-刘畅
> Dear @duncanriach, > > I'm using TF2.3 on Ubuntu 16.04. > To get deterministic results, I followed your instructions. Please check the attached code that is very simple MNIST...
> DepthwiseConv2D Not yet. Listed below are codes for network copied from [ResNet_TF2](https://github.com/Apm5/ImageNet_ResNet_Tensorflow2.0) ``` category_num = 1000 block_type = {18: 'basic block', 34: 'basic block', 50: 'bottlenect block', 101: 'bottlenect...
请问你解决问题了吗？ 求提供 + 1
I have the training code but not all files for training. I haven't reproduce it yet ```python #!/usr/bin/env python # Copyright 2016 Google Inc. # # Licensed under the Apache...
Additional, I use cv2.PSNR.
**I changed my code to compute psnr.** > In OpenCV:   PSNR = 10 * log10( range / mse of all 3 channels ) In your codes:   PSNR = Mean( 10...
我觉得不太可能。因为双边网络上采样的部分是用 C++ 写的，里面要 cuda，所以要 CPU 版本只能自己写
目前不能用 1.7.+
> 计算Grad-CAM的时候应该有一点问题，权重是根据梯度计算的，而不是原始特征图 > > ```c++ > data_type *const fea_ptr = delta[b]->data + o * area; > ``` 你好！谢谢指正，本人会抽空看看论文和具体实现，谢谢！
Some extra nms should be done after nms on every class : **In the https://github.com/leviome/yolo_1_pytorch/blob/e18705cbf22034d97e043e5faa22466c129238d5/yolo/decoder.py#L57**  And the results are more accurate as listed below:   