S.EB

1

votes
1

answer
740

views

How to train a FCN network while the size of images are not fixed and they are varying?

I have already trained the FCN model with fixed size images 256x256. Could I ask from experts how can I train the same model once the size of image are changing from one image to another image? I really appreciate your advice. Thanks
S.EB
1

votes
0

answer
96

views

How to load a multi-channel (5 channel) into FCN which has been already trained on RGB images?

I am using FCN8s and its pre-trained model for semantic segmentation on my data. Since vanilla-fcn models accept three channels images (RGB), I do not know how can I give 5 channel images to the model. Could someone please explain how can I give an input to a model that is using a pre-trained model?...
S.EB
1

votes
1

answer
709

views

Why I am getting this error: object has no attribute 'astype'

I have defined a function myfunc inside a python file myfile. when i import this function from a jupyter notebook, it is showing the following error: import numpy as np import os from scipy.misc import imread import ast from myfile import myfunc ..... class_mask = np.equal(image, i) class_mask = cla...
S.EB
1

votes
0

answer
178

views

ImportError: No module named vlfeat

I have installed VLFeat on anaconda environment successfully, but once I wanna import it, shows >>> import vlfeat Traceback (most recent call last): File '', line 1, in ImportError: No module named vlfeat Any idea about this? thanks
S.EB
1

votes
1

answer
570

views

Why loss is remaining constant during training FCN-8s?

I am trying to run FCN-8s. I did the following steps: 1. downloaded this repository 2. converting my data to LMDB and changing the paths in train_val.prototxt 3. downloading the fcn8s-heavy-pascal caffemodel 4. changing the number_of_output in train_val.prototxt and deploy.prototxt from 60 to 5 (th...
S.EB
1

votes
4

answer
2.9k

views

How can I plot the log file?

Does anyone know any function for plotting the obtained measures in Caffe? I would like to plot train loss, test loss, and accuracy, train moving average and etc. in one plot. Is there any function except Caffe built-in function that is available online? Edited: First, I ran parse_log.py file (the f...
S.EB
1

votes
1

answer
379

views

How to solve this issue in Python (creating weights for Infogain Loss layer)?

I am working on semantic segmentation using CNNs. I have an imbalance number of pixels for each class. Based on this link, I am trying to create weight matrix H in order to define Infogain loss layer for my imbalance class members. My data has five classes. I wrote the following code in python: Rea...
S.EB
1

votes
2

answer
6.6k

views

ImportError with importing keras

I am trying to import keras it gives me an error from tensorflow.python.ops import tensor_array_ops. >>> import keras Using TensorFlow backend. Traceback (most recent call last): File '', line 1, in File '/home/ss/anaconda2/envs/kerasenv/lib/python2.7/site-packages/keras/__init__.py', line 2, in...
S.EB
1

votes
2

answer
59

views

I am getting a black image in FCN32

I trained FCN32 from the scratch on my data, unfortunately I am getting a black image as output. Here is the loss curve. I am not sure whether this training loss curve is normal or not, or whether I have done something wrong or not. I really appreciate experts'idea on this. And why the output is a...
S.EB
1

votes
2

answer
297

views

I am getting error with im.show()

I am trying to save a gray scale image (256,256,1) and show it in the output. im = data.astype(np.uint8) print im.shape im = np.transpose(im, (2,1,0)) print im.shape im.show() However, I am getting the following error: (256, 256, 1) Traceback (most recent call last): File 'lmdb_reader.py', line 37...
S.EB
1

votes
1

answer
66

views

What Kind of pre-processing techniques can I apply to make an object more clear?

I applied few techniques of denoising on MRI images and could not realize what techniques are applicable on my data to make the cartilage object more clear. First I applied Contrast-limited adaptive histogram equalization (CLAHE) with this function: J = adapthisteq(I) But I got a white image. This i...
S.EB
1

votes
1

answer
24

views

Is it possible to use pretrained model after adding elementwise layers?

I am using a pre-trained model which I want to add Elementwise layer that products the output of two layers: one layer is output of convolution layer 1x1x256x256 and the other is also the output of convolution layer 1x32x256x256. My question is: If we add elementwise layer for multiplying two layers...
S.EB
1

votes
1

answer
60

views

issue of including QtGui files in building a c++ based software

I am not expert in c++ and I am trying to install a software which is based on c++; during the building the files in Qt Creator, once it reaches to this line to include #include #include it shows error /home/ubuntu/Ezys/source/Ezys_GUI/ezys_gui.h:12: error: QtGui/QMainWindow: No such file or dire...
S.EB
1

votes
1

answer
4k

views

How to remove one of the protobuf compiler that is making conflict?

I have installed protobuf using this conda install protobuf==2.6.1 in a conda environment. So once I want to install a software it showing the following error message: #error This file was generated by a newer version of protoc which is #error This file was generated by a newer version of protoc whi...
S.EB
1

votes
1

answer
181

views

How to use FLANN for labeling and clustering?

I read a paper that their retrieval system is based on SIFT descriptor and fast approximate k-means clustering. I installed pyflann. If I am not mistaken the following commands only find the indices of the close datapoints to a specific sample (for example, here, the indices of 5 nearest points from...
S.EB
1

votes
1

answer
82

views

How to find the matched SIFT features that are spatially consistent?

I have extracted DenseSIFT from the query and database image and quantized by kmeans using VLFeat. The challenge is to find those SIFT features that quantized to the same visual words and be spatially consistent (have a similar position to object centers). I have tried few techniques: using FLANN()...
S.EB
1

votes
1

answer
0

views

Setting specific values to 0 in a NumPy array

I have two arrays (gt and pred) with values ranging from 0 to 4. The shape of these two arrays is (1, 1, 93, 349, 219). My target is to create a mask to ignore and multiplying it into gt and pred to ignore the value in two arrays. However, I am facing an issue ignore_value=4 if ignore_value is not...
S.EB
1

votes
1

answer
0

views

How to convert a specific range of elements in a panda DataFrame into float numbers?

I have a panda Dataframe like the following and this is the data: 0 1 2 3 4 5 6 0 Label Total/Target Jaccard Dice VolumeSimilarity FalseNegative FalsePositive 1 image-9003406 0.753958942196244 0.628584809743865 0...
S.EB
0

votes
0

answer
4

views

Finite element analysis: How can I find the boundary conditions and settings for the knee joints?

I am not an expert in finite element models (FEMs), however, I am reading related articles and trying to find the boundary conditions for knee joint including the bone and cartilage. Does anyone have any suggestion? How can I find the boundary conditions for the knee joint? If there any platform or...
S.EB
0

votes
1

answer
21

views

IndexError: 'index out of bound'

I have 2 arrays one is mask and the other is the labels: Both arrays have the same shape: (Pdb) L.shape (178, 201, 101) (Pdb) MASK.shape (178, 201, 101) when it reaches to this line: L[~MASK] = 0 IndexError: 'index 255 is out of bounds for axis 0 with size 178' it shows an error that I could not fi...
S.EB
1

votes
1

answer
1.4k

views

Is there any practical tutorial for Conditional random fields (CRF) and Markov random fields (MRF)?

I have started reading on MRF and CRF in a short-term duration; however, I have a lot of difficulty in understanding the concepts. Can anyone suggest some online resources for theoretical and basic understanding? I have a lot of difficulty in understanding maths of these two concepts. Besides, I wou...
S.EB
1

votes
1

answer
629

views

Getting error on [base_conv_layer.cpp:122] Check failed: channels_ % group_ == 0 (1 vs. 0) , how to solve it?

When I am trying to train FCN32 for semantic segmentation on my own data, I am getting this error: I0106 12:57:53.273977 19825 net.cpp:100] Creating Layer upscore_sign I0106 12:57:53.273982 19825 net.cpp:434] upscore_sign upscore_sign F0106 12:57:53.274119 19825 base_conv_layer.cpp:122] Check fail...
S.EB
1

votes
1

answer
427

views

How to apply a pre-trained model of 3 channel images on single channel images?

I tried to used a pre-trained model that already was trained on three-channel color images, however, I am getting an error because of shape difference. Could someone let me know how can I tackle this issue? One user had suggested using Tile layer, but I could not find any relevant document/help for...
S.EB
1

votes
1

answer
251

views

How can I create a fresh (without any package installed) conda environment?

I am installing a software which needs some pre-requisite packages in a conda environment. There are some incompatibilities with the packages that I have already installed on anaconda and I need to install these packages on a fresh environment without any packages installed on that. I created conda...
S.EB
1

votes
1

answer
356

views

Is there any part of speech tagger and tokenizer of Tamil language?

I am a beginner in natural language processing. I have to work on different languages that Tamil is one of them. Could I ask from experts whether there is any Tamil language tokenizer code (java,c,python or etc.) and part of speech tagger codes that I use it for my research? I really appreciate if...
S.EB
1

votes
1

answer
2.5k

views

cp: not writing through dangling symlink '/usr/lib/x86_64-linux-gnu/libhdf5.so'

I am facing this issue, Could you please give me some hints how to solve this: $sudo cp ~/anaconda2/lib/libhdf5* /usr/lib/x86_64-linux-gnu/ cp: not writing through dangling symlink '/usr/lib/x86_64-linux-gnu/libhdf5.so' cp: not writing through dangling symlink '/usr/lib/x86_64-linux-gnu/libhdf5_hl.s...
S.EB
2

votes
1

answer
623

views

What parts an parameters of train_val.prototxt, solver.prototxt, and deploy.prototxt should be changed based on our own data?

I am confused about changing the parameters for training a net in caffe based on our own data. Which layers of net we should pay attention more to train the net on our own data? For example, the number of outputs based on the number of classes. I tried to train FCN32 for semantic segmentation. I ch...
S.EB
3

votes
1

answer
644

views

How to design a CNN (the inputs and outputs of the layers)?

I am a beginner and learning deep learning with baby steps. I have a question about designing the nets. I see in the papers, there are layers with different inputs/outputs and I do not know how to calculate/design before implementation. For instance, in this paper, there are some numbers beside the...
S.EB
2

votes
1

answer
420

views

I am getting Incorrect data field size error while generating the mean image of training data, even after generating by using create_imagenet.sh,why?

I am trying to generate the mean image from training data in caffe. My data is 256x256 grayscale images. I created lmdb by using create_imagenet.sh by replecing --shuffle with --gray. I edited create_imagenet.sh as follows: GLOG_logtostderr=1 $TOOLS/convert_imageset \ --resize_height=$RESIZE_HEIGHT...
S.EB
2

votes
1

answer
253

views

What is the purpose of dropout layer in deploy.prototxt?

I read somewhere, during testing there is no Dropout applied. What is the purpose of dropout layers in deploy.prototxt? Is not deploy file for testing the model? Thanks in advance
S.EB
1

votes
1

answer
33

views

Finding object center: showing wrong coordinate outside of the target object

I followed the code in this link to find the object center for my graylevel image, def find_center(im): immat = im (X, Y) = [im.shape[0],im.shape[1]] m = np.zeros((X, Y)) for x in range(X): for y in range(Y): m[x, y] = immat[(x, y)] != 0 m = m / np.sum(np.sum(m)) # marginal distributions dx = np.sum...
S.EB
2

votes
1

answer
314

views

How can I save the probability of the prediction in caffe?

Does anyone know how can I save the output of predicted class labels of each pixel in FCN semantic segmentation? I would like to see the probability map of the image during inference. The data in which layer should be saved? Many thanks
S.EB
2

votes
1

answer
523

views

How did you run FCN code semantic segmentation?

I wanted to run the [FCN code][1] for semantic segmentation. However, I am beginner in Caffe and I did not know from which point should I start running the code. Is there any step by step guidance for running?
S.EB
2

votes
0

answer
277

views

Error with Caffe installation (cmake command): LINK_INTERFACE_LIBRARIES

Hi when I want to run cmake for caffe installation it is showing me this: Target 'caffe' has an INTERFACE_LINK_LIBRARIES property which differs from its LINK_INTERFACE_LIBRARIES properties. INTERFACE_LINK_LIBRARIES: proto;/usr/lib/x86_64-linux-gnu/libboost_system.so;/usr/lib/x86_64-linux-gnu/libboos...
S.EB
2

votes
1

answer
537

views

Tackling imbalanced class members in Caffe: weight contribution of each instance to loss value

I have a highly imbalanced data, I know that some users suggesting using InfoGainLoss loss function, however, I am facing few errors when I tried to add this function to Caffe layers. I have the following questions, I really appreciate if someone guides me: How can I add this layer to Caffe? Does an...
S.EB
2

votes
1

answer
49

views

How can I start writing the code for my layer?

I have seen that researchers are adding some functionalities to the original version of Caffe and use those layers and functionalities according to what they need and then these versions are shared through Github. If I am not mistaken, there are two ways: 1) by recompiling Caffe after adding c++ and...
S.EB
4

votes
1

answer
495

views

Is there any example of using weighted loss for pixel-wise segmentation/classification tasks?

I am doing FCN32 semantic segmentation on my data. I ran the algorithm to fine-tune for my data (grayscale images with only one channel), till 80,000 iterations; however, the loss and accuracy are fluctuating and the output image completely black. Even, the loss is so high after 80,000 iterations. I...
S.EB
2

votes
1

answer
153

views

Probability as input to Markov random field (MRF): how to refine the cmex code?

I am very new with MRF and not that much good at programming. I have obtained probability map from semantic segmentation using a CNN, I have to optimize the segmentation by using Markov Random Fields (MRF). I download the code provided by Shai Bagon in this link GCmex. Energy minimization is perfo...
S.EB
0

votes
0

answer
133

views

Generalized hough transform for center of an object calculation based of SIFT points?

Do You have any suggestion for this? I am following the steps in one research paper to re-implement it for my specific problem. I am not a professional programmer, however, I am struggling a lot, more than one month.When I am reaching to scoring step with generalized Hough transform, I do not get an...
S.EB
3

votes
1

answer
56

views

How to define a Gaussian voting for generalized hough transform within a range of displacement of a coordinate in accumulator?

I am not professional programmer and taking baby steps in that. I have defined a Gaussian kernel which shows the weights for scoring with a window size of 5 and sigma=3 (i.e., 3 displacements) in python. [[0.00364954 0.01483356 0.02344521 0.01483356 0.00364954] [0.01483356 0.06029105 0.09529314 0.0...
S.EB

View additional