Questions tagged [image-segmentation]
The image-segmentation tag has no usage guidance.
74
questions
0
votes
0answers
7 views
Can Dice Similarity index be higher than Jaccard index (Mean IOU) for a segmentation task? If so why does this happen?
I am currently doing a segmentation task. DICE is 0.94 and Mean IOU (Jaccard Index) is 0.63. How can mean IOU be lesser than DICE and is this possible. Also need explanations on scenarios where ...
0
votes
1answer
12 views
Hook up PyTorch U-Net model to video
I built a U-Net model in PyTorch that is trained on medical images to detect polyps. The purpose of the model is to do semantic segmentation, so it must predict the location + class of polyps.
Now I ...
0
votes
0answers
11 views
Detection of vortex and numbering on basis of rotation
Problem Statement: I have a video of ANSYS Simulation of vortices formed due to flat plate plunging. The video contains vortices (in simpler terms blobs), which are distinguished according to their ...
0
votes
0answers
16 views
Computing symmetric difference hypothesis divergence $H \Delta H$ for two domains using a segmentation network
Given two domains $D_1$ and $D_2$, the symmetric difference hypothesis divergence ($H \Delta H$) is used as a measure how much two domains differ from each other.
Let the hypothesis, segmentation ...
1
vote
1answer
26 views
How to design a model for contour recognition? In particular, how to shape the output layer?
I want to design and train a neural network for the automatic recognition of the edges, in some microscopic images.
I am using Keras for a start, I may consider PyTorch later.
The structure of the ...
0
votes
0answers
7 views
How to check if 2 images where splitted?
There is an original data-set of images. Each image was splitted into 2 parts (left-right).
I want to run on all those splitted images and check if each 2 images are spliited from same image.
Is ...
2
votes
0answers
21 views
Instance Segmentation using the predefined bounding boxes
I want to do Instance Segmentation using the images in my dataset which are already annotated and I don't want to train the model but use the pre-trained model. I was following this colab notebook. ...
2
votes
1answer
18 views
Segmentation model to predict face forward and profile parts of the face
I am developing a model for feature counting on a person's face that consumes three photos (one face forward and two profile pictures). My model can already detect features, but it counts some ...
1
vote
1answer
26 views
How to resize image along with their mask?
I have original images of the size 1935x1481. I am using labelme to annotate the images. I am creating polygons on the original image. Is there a way to resize the image along with their mask? I am ...
1
vote
1answer
15 views
Why do we need to concatenate in a U-Net?
You might be familiar with the U-Net, a machine learning network deceived for image segmentation.
It's basically an encoder/decoder network with some direct links between encoder and decoder segments:
...
1
vote
0answers
23 views
Beginner level: how to interpret LIME and classification result
I am new to the concept of model interpretability using LIME method. I am following the tutorial LIME for spectrogram classification. I am finding hard to understand the color coding -- before using ...
2
votes
0answers
25 views
Transfer Deep Learning from one aerial imagery datset to many others
I am new to Deep Learning but have been able to use RasterVision successfully to predict building footprints within a set of aerial imagery.
This aerial imagery data set is for a province of New ...
1
vote
1answer
41 views
Correct way of computing dice score for image segmentation?
In binary image segmentation, for given a set of images, it's true mask and predicted mask. How do you compute dice score? Should I compute the dice score for each image separately and then find mean ...
0
votes
0answers
29 views
How to use mean IoU for RGB mask (keras implementation)?
I am training pix2pix GAN for converting SAR satellite images to segmentation mask. But I am not aware about how to use the mean IoU to evaluate my model. My output is of ...
0
votes
0answers
52 views
Best algorithm for real-time instance segmentation in videos?
I want to do object detection in real-time (meaning localization and classification) on videos (25 FPS), but with the added constraint that my training data is labelled using binary masks, rather than ...
1
vote
1answer
22 views
3 images as one input in CNN (U-Net) [closed]
I have been advised by my supervisor that if my U-Net segmentation network has RGB images at the input then I could use the channels for different images - median filter for R, normalization for G, ...
1
vote
1answer
182 views
How can I create .nii (nifti) file from 3D Numpy array
I have a prediction numpy array. How can I make a .nii or .nii.gz mask file from the array?
1
vote
1answer
63 views
Modifying U-Net implementation for smaller image size
I'm implementing the U-Net model per the published paper here. This is my model so far:
...
0
votes
0answers
16 views
U-Net doesn't work with images different from the dataset
I have implemented a very similar U-Net code from github, but for a different dataset, this one, to segment roads, it works fine using the test folder images, but when i for example, pick a print from ...
1
vote
1answer
29 views
Implementing U-Net segmentation model without padding
I'm trying to implement the U-Net CNN as per the published paper here.
I've followed the paper architecture as closely as possible but I'm hitting an error when trying to carry out the first ...
0
votes
0answers
9 views
Instance segmentation mask creation for adjacent objects within the same class
How are instance segmentation masks created for adjacent objects from the same class ? If for example we have different objects, we can distinguish between them using the labels, but how to ...
0
votes
0answers
7 views
Boundary segmentation
I have one problem regarding segmentation. I would like to put more concentration on the boundary instead of the interior of segmented part. Is that possible to do using Tensorflow 2? For example, if ...
0
votes
1answer
24 views
Semantic segmentation in high-resolution images with high variance - cannot avoid underfitting
I am working on a dataset of 2K images for a semantic segmentation problem.
I want to detect and localize small objects, with the smallest mask to be 5x5 pixels.
The images include 5 different ...
1
vote
1answer
126 views
Creating parallel keras layers
I am new to Keras and ML and I want to create a NN that can seperate a bitmap-like image into its visual components.
My approach is to feed a two dimensional image (lets say 8 by 8 pixels) into a NN, ...
1
vote
1answer
19 views
Photorealistic synthetic data for object segmentation
Let's say that we have a very small labelled dataset for instance segmentation and there is a photorealistic physics engine available that can produce synthetic data for us. Looking on the web I haven'...
0
votes
0answers
19 views
Changing order of input dimension in Tensorflow 3-D Layers
According to the official documentation of tf.keras.layers.Conv3D
5+D tensor with shape: batch_shape + (channels, conv_dim1, conv_dim2,
conv_dim3) if data_format='channels_first' or 5+D tensor with ...
1
vote
1answer
88 views
What is Deep learning approach to count the number of Diamonds in an image?
I am working on a project which involves counting the number of diamonds in the provided image. I have a set of images and a VIA annotated .json file which has all the annotations. How do I proceed ...
0
votes
0answers
36 views
Agglomerative Hierarchical Clustering on Images
My goal is to implement the agglomerative hierarchical clustering algorithm on an RGB image to cluster every pixel until some stopping criteria is reached. In order to do so, I assumed that each pixel ...
0
votes
0answers
13 views
Semantic segmentation to instance segmentation
Having a mask from semantic segmentation I want to split it to list of non-overlapping instances masks(like an output from instance segmentation). Do you know some fast algorithm, approach, tool or ...
1
vote
1answer
25 views
What do the parameters used in crop mean?
When we have an image to be used as an input to a CNN and we want to classify only part of the image, we usually feed the classifier with a crop of the image.
Lets say my image is called frame and <...
0
votes
0answers
11 views
I have no access to gpu due to usage limits?
I start running my code using google colab I first set the execution to GPU
and then I run my code for a training task using keras !after 1 hour I got a message saying I can't use GPU due to usage ...
0
votes
0answers
155 views
Implementing Multiclass Dice Loss Function
I am doing multi class segmentation using UNet. My input to the model is HxWxC and my output is,
...
0
votes
0answers
14 views
Satellite Change Segmentation using Unet
Hi StackExchange community I am working on to train a Unet for satellite change segmentation. My dataset consists of images(before change),images(after change) and the corresponding change ...
0
votes
0answers
11 views
How to generate fixed number of superpixels?
A lot of work regarding Graph Neural Networks require fixed number of nodes. In the case of image processing using graph, the image representation is often super pixels (like in this work https://...
1
vote
0answers
99 views
How do you deal with variable input sizes with an encoder-decoder net with skip connections in Keras?
I am currently getting into image segmentation with Keras, and I am using an encoder-decoder type as in the image below.
My problem is that applying a MaxPooling2D ...
0
votes
0answers
17 views
Loss function to compare non binary segmentation
I need to compare two images corresponding to landmark locations. I was thinking of something related to Dice loss. I cannot use dice loss since the image is not binary. The background is black but ...
3
votes
1answer
90 views
Tool for annotation of images for semantic segmentation
I have been searching around for a software tool, that I can use for annotating images. More specifically I want to do annotation to be used for semantic segmentation, meaning I want to create masks. ...
1
vote
1answer
194 views
Normalization of CT scans
I trained an infection segmentation models on a large dataset of CT scans, and want to extend it to other datasets to show the ability of the model to generalize. What I found though, is that CT scans ...
1
vote
1answer
38 views
Ways of calculating the area of colored regions in a map
Background
I am a PHD student trying to improve my data science. One of my research projects, has me tasked with determining the size of the clusters in a colored image of regions. Here is an example ...
0
votes
0answers
26 views
loss function for multi-label segmentation with class inbalance
In order to use a binary segmentation loss function in a multi label problem, I would like to permute the batch axis with the channel axis in the loss computation in order to compute the loss by ...
0
votes
0answers
11 views
How to Convolve a High-Res Image by a (fully convolutional) CNN kernel?
My CNN is an extremely simple neural network.
...
0
votes
0answers
26 views
IOU and dice_coef exceed 100% in multi task learning
Im doing a multi task learning for road and center line extraction (2 classes) I used IOU and dice_coef as a metrics :
...
1
vote
1answer
108 views
Extremely stochastic validation loss/accuracy
I am working on training DNNs on satellite data. The class distribution in the data is extremely imbalanced, so I train the neural networks using random majority undersampling to artificially balance ...
1
vote
1answer
463 views
How can I use my own dataset for Image segmentation using Tensorflow
I have a huge problem using my own created dataset for image segmentation using Tensorflow. The dataset that I've build contain images like the one shown below:
The problem that I have is: How do I ...
0
votes
1answer
63 views
For semantic sementation, why am I getting better loss values with binary cross entropy than dice coef?
I'm learning all related to data science and how to train U-Net to do semantic segmentation.
I have a U-NET with this loss function:
...
0
votes
0answers
20 views
Model selection Tensorflow for custom dataset comparing panoramic images vs regular images (Image segmentation)
I have a question regarding the use of a specific model such as Deeplab and how to create a custom dataset for it.
Background
To give a bit of background info, I want to compare panoramically stitched ...
0
votes
0answers
18 views
Why it is reshaped the last layers of VGG_UNet segmentation model?
I want to do a multiclass segmentation task using deep learning (in python). Here, is a summary of vgg_unet model that is mainly collected from GitHub. So, in my dataset 8 labels are available. So, at ...
0
votes
0answers
16 views
Is it possible to create a 3d to 2d U-net?
I was curious if it is possible to create a U-net type architecture that takes in a 3d image and outputs a 2d image? Or, alternatively, would some other architecture be better suited for this problem? ...
0
votes
1answer
10 views
Which colour channel from a TIFF image do I have to use?
I'm going to use the following dataset to do semantic segmentation with U-Net network.
LGG Segmentation Dataset
This dataset contains brain MR images together with manual FLAIR abnormality ...
1
vote
1answer
107 views
Multiple output size in neural network
In the paper "A NOVEL FOCAL TVERSKY LOSS FUNCTION WITH IMPROVED ATTENTIONU-NETFOR LESION SEGMENTATION" the author use deep supervision by outputing multiple outputmask which have different ...
