wangq95

About remapping prediction to colorful maps.

1

Hi, I am confused about that how to get the colorful prediction like that:  Could you give me some advice about that? Best regards!

Hi, I once used this script to test the speed of ICNet in Matlab : tic; score = net.forward({input_data}); temp = toc; But I got the result of 15.6fps on...

The prototxt files for training PSPNet is not available, can you provide the code for training? Thanks.

Hi, @ansleliu , thanks for your work, can you share the training protocal, **lr, batchsize, optimizer, epochs etc.**, for training on Cityscapes? Thanks a lot.

Could not run the code

1

Hi, @speedinghzl , I try to train on Cityscapes dataset using pytorch 0.4.0, but I got an error as follows: Traceback (most recent call last): File "train.py", line 253, in...

Error in dual task loss?

1

Hi, @tovacinni Thanks for your work, I have a question about the implementation of dual task loss in loss.py. I think the dual task loss should take the estimation from...

Hi, @MarvinTeichmann Thanks for your work, I have two questions about the ConvCRF. 1) I run demo.py and got a perfect result, but how the weights of Gaussian Kernels been...

Hi, when I try to finetune the model after 250000 iterations, but I got an nan loss without any ideas, can you help me to solve this problem? Best regards!...

Training is too slow

2

Hi, @Tramac , I use your code with default setting, and find the speed of training is too slow. The usage of GPU is less than 10%, resulting in about...

Error information: In instantiation of ‘void caffe::ContrastiveLossLayer::Forward_cpu(const std::vector&, const std::vector&) [with Dtype = float]’ Open src/caffe/layers/contrastive_loss_layer.cpp Replace the 56th line with: Dtype dist = std::max(margin - (float)sqrt(dist_sq_.cpu_data()[i]), Dtype(0.0)); Save and...