Kai Li

The loss of age does not decrease?

1

The loss of age does not decrease, about equal 3.6

why the input is 112X96 ?

1

How did you choose this number? why not squares? Thank you very much.

Lasagne/Theano model running much slower than the same model using Caffe

5

I used to use Caffe for my research. Now I want to use Theano based framework since it is much more flexible than Caffe. I choose to use Lasagne now,...

Best practice for training on large scale datasets like ImageNet using Lasagne

9

Hey guys, I found that all of the examples of Lasagne deal with small data set like mnist and cifar10 which can be loaded into memory completely. My question is...

how to extract key frames of a video?

5

My question is How to use lintel to get the same results of the following ffmpeg command `` ffmpeg -v quiet -skip_frame nokey -i video.mp4 -vsync passthrough img_%04d.jpg `` Thanks

The speed of Deep_Residual_Learning_CIFAR-10.py

16

I am using Ubuntu 14.04 with a titan x GPU and cudnn v4. When I run the Deep_Residual_Learning_CIFAR-10.py script The results are as follows ``` Using gpu device 3: GeForce...

RandomIdentitySampler

1

https://github.com/xiaolai-sqlai/gluon-reid/blob/master/branch/data_read/sampler.py#L35 这里的len是不是应该返回self.num_samples？也就是_iter_函数中ret的长度 我看https://mxnet.incubator.apache.org/_modules/mxnet/gluon/data/sampler.html#BatchSampler batchsampler的源码返回的是batch的数目

compile error

1

I copy my own Makefile.configure to your caffe-public, Then make -j10 ``` AR -o .build_release/lib/libcaffe.a LD -o .build_release/lib/libcaffe.so CXX/LD -o .build_release/tools/compute_image_mean.bin CXX/LD -o .build_release/tools/extract_features_binary_output.bin CXX/LD -o .build_release/tools/test_net.bin CXX/LD -o .build_release/tools/finetune_net_match.bin...