Questions tagged [computer-vision]

Questions related to image representation, segmentation, visual object categorization and image processing algorithms in general.

Filter by
Sorted by
Tagged with
0
votes
0answers
20 views

How to transform from interal of p(z)dz to interal of p_g(x)dx in GANs

According to mathematics, I don't understand why the interal of p(z)dz can change into the interal of p_g(x)dx. I know that G(z) = x. But G() is a function which I don't know -> can't manually ...
0
votes
0answers
23 views

How is it possible that validation loss is changing but validation accuracy repeating the accuracy pattern?

I'm using the pre-trained model of MobileNet. By making use of transfer learning, I'm training the last 25% of layers of this architecture. It's a binary classification problem. Using Keras for this. ...
0
votes
0answers
5 views

Will running many classification models sequentially be similar to running only 1 detection model?

This question is related to this other one: In what situation many classification models (specialized) will be better to use instead of just 1 (generalist)? With 1 classification model it is possible ...
0
votes
0answers
12 views

In what situation many classification models (specialized) will be better to use instead of just 1 (generalist)?

Recently, I started working on Classification and Object detection. I am using a dev board to make ONLY inference. So I am just using the models created by someone else. As I have to develop a demo, ...
0
votes
0answers
6 views

How Milestone object detectors differ from each other?

I'm tracing the object detector types starting from traditional detectors like Viola Jones, then machine learning detectors and ending with deep learning by Yolov4. Therefore I've created a table that ...
0
votes
0answers
21 views

Train a neural network to find position of altered pixels?

I have a series of images, all of which have been altered in the same way. A circle of radius, r, containing a face or other object has been edited out of the pictures and replaced with a best attempt ...
0
votes
0answers
5 views

Acceptable level of mAP in computer vision applied to health applications

I am trainig convolutional methods for detecting dental features in radiographs and trying to get the highest mAP by fine tuning, training new models and improving the ground truth labels by paying ...
0
votes
0answers
16 views

Components and Purposes of CNN Architecture

I am trying to learn about Convolutional Neural Networks. I have begun to study the components of the architecture and I think I have a good idea of the primary components that make up most ...
0
votes
0answers
20 views

What is 'prototype learning' in computer vision?

I was reading some papers about semantic and instance segmentation (e.g. YOLACT) and often encountered the term 'prototype learning'. Could you please explain this concept?
0
votes
1answer
16 views

How can I infer the cost function from Kruppa's simplified equations?

The following equations are Kruppa's simplified equations used in camera autocalibration. My objective here is to infer the cost function(Error Function) from this equations, So I can minimize the ...
0
votes
1answer
30 views

why predict a distribution in pixelcnn++ [closed]

I know that in the original pixelcnn paper, they predicted a 255 vector for each subpixel, and argmaxed to get the value. in the pixelcnn++ paper, if I understand it correctly, they model the pixel ...
0
votes
0answers
7 views

Finding Patterns on Surface of Material

I recently got a set of images on which I need to find certain patterns (such as scratches or roughness). Now I have two types of surfaces. I would like to compare if these two surfaces exhibit the ...
0
votes
1answer
21 views

How to compute feature-space distance?

some CV research papers use nearest neighbors techniques to compare images, such as PROGRESSIVE GROWING OF GANS FOR IMPROVED QUALITY, STABILITY, AND VARIATION Next five rows: Nearest neighbors ...
0
votes
0answers
21 views

Non-linearity (ReLU, Batch Norm) before final sigmoid convolution in image segmentation

When building an image segmentation model, standard building blocks consist of Conv -> RELU -> Batch Norm -> DropBlock. I understand the importance of having a non linear activation, and the role that ...
0
votes
0answers
24 views

Intersection vs Chi-Square for comparing histograms

I've seen lots of examples where chi-square works better. But is there any example of for ex. 3 non-similar histograms where Intersection results in the similarity of histograms but Chi-square results ...
0
votes
0answers
7 views

Techniques for estimating homography

Is there any good comparison of methods (SIFT+RANSAC, SURF, MSER, ConvNets...) for estimating homographies / finding correspondencies? Or, if you have experience in this topic, what would be the best (...
0
votes
0answers
10 views

Annotation of dataset for object detection

I am annotating a dataset for object detection of UI elements (buttons, text, edit text, etc..) in images of phone screens and I am wondering what is a better approach. If I want to detect buttons (...
0
votes
1answer
33 views

How to develop a Deep learning model for only two Imges …?

I am trying to build Deep learning model for only two Images. I have a two images of doors , one image is some what good and another is some what bad, I want to identify the one is good and another ...
0
votes
0answers
36 views

Is object detection the right approach for this problem

I'm trying to build a model which, given a picture of someone's face, is able to identify all the following features, as well as others. The model would output, for each picture, a list of all the ...
0
votes
0answers
18 views

Computer vision with smaller data set

I hope this is the right forum to ask. I had a client approach me with a demand for a vision system for their assembly line. The problem they are facing is that the operator sometimes forgets to put ...
0
votes
1answer
29 views

How to use additional non-visual data for image classification?

I've got a Resnet network that classifies images into n classes and is working fine. I want to boost its performance buy using additional information I have regarding the images. This info comes in ...
0
votes
0answers
106 views

Reading a pressure gauge with a CNN

Using standard Computer Vision pipelines to read pressure gauges is well established, and not overly accurate or generalizable: For various reasons, I would like to use a CNN to do this. Since the ...
0
votes
0answers
81 views

Retraining of object detection CNN

I am working on an object detection system that should detect UI elements (such as button, checkbox, radio button, etc..) in the photo of a touch screen of printer (not screenshots, but literally a ...
0
votes
0answers
18 views

video normalization using skvideo

I'm building a model that would take as input a video of 25 frames and would (ideally) output the next 25 frames. My question is when we use images we usually normalize by dividing the X by 255. ...
0
votes
0answers
14 views

Image classifier from text files

I have a data set of thousands of images of hundreds of pixels in gray scale ranging from -1 to 1. The labels represent 0 to 9. The issue is that the data set is in .txt format. How can one make an ...
0
votes
1answer
61 views

Object detection : Multimodal or single input ? for Depth + Thermal images

I need to detect persons in a scene. I have a 16 bits depth image of that scene (640, 480) I have a 16 bits thermal image (80, 60) of the same scene (slighly different point of view) I resize these ...
0
votes
1answer
101 views

Is there a difference between training with multiple objects in a single image and multiple objects in a different images?

I'm trying to generate data for my object detection network (which will be used for TensorFlow: ResNet). What I'm currently curious about is this: if I have the same total amount of data (each data ...
0
votes
0answers
40 views

Is there a model or algorithm to improve digital drawings?

given a bad drawing, the algorithm should deform the edges of the bad drawing and fill them with color so that it looks more like a learned character. the entry would be a very poorly made drawing of ...
0
votes
1answer
55 views

Is it possible to train a model to detect a shape? [closed]

We have multiple object detection APIs that can help us to find the bounding box coordinates. But is it possible to go a little further and find/separate the shape of the object (say cat/dog/human) ...
0
votes
0answers
22 views

What do I do when my deep learning model in CNN does not learn? [duplicate]

I'm training on a skewed image dataset of 5000 images with class weights. The training loss decreases well and swift, but the validation-set loss fixates after around 6 epochs. I have tried using my ...
0
votes
1answer
120 views

Object distortion after ROI Align in Mask R-CNN

In Mask R-CNN, if there are 2 proposed ROIs which cover 2 objects that looks like below: #1 A square object #2 A rectangular object So my question is: After ROI Align, is the #2 feature map ...
0
votes
0answers
176 views

How to reproduce the classification heatmap figure (Figure 2) from the FCN paper?

I want to replicate the results of the paper "Fully Convolutional Networks for Semantic Segmentation" (Long and Shelhamer, 2015) in PyTorch. Doing so, I tried to replicate the initial experiment ...
0
votes
1answer
79 views

How to deal with different image sizes during training and inference? (e.g. Stacked Hourglass)

In some of computer vision papers I read that they start off with a bigger sized image and use pooling to reduce dimensionality and train on the image with lower resolution. However, they don't ...
0
votes
0answers
43 views

Unsupervised classification of images

Assuming I have a dataset of images from two similar classes, for example let's say 95% Labradors and 5% of Chihuahuas and I want to make a classifier. The point is that I need to find the anomalies (...
0
votes
1answer
16 views

IOU vs ROBIN metrics

Came across this paper on ROBIN evaluation metrics. The metrics seem to be more informative than just IOU, so is there a reason why IOU is the preferred metric in most cases for object detection. ...
0
votes
0answers
25 views

what is the best approach in dealing with large dimension custom data for training and predicting deep learning models

i am trying to implement semantic segmentation for satellite images.My custom dataset has dimensions(height,width)in range (3000, 3000)what is the best approach for feeding(for training) and ...
0
votes
0answers
58 views

How to separate training set and validation set using 80/20 rule?

I have two folders: One folder contains images of non-dogs, and the second folder contains images of dogs. I am to divide these folders into a "training set" and a "validation set" with the 80/20 rule....
0
votes
1answer
48 views

Why do we use Hough instead of RANSAC in SIFT?

In his SIFT paper, why did Lowe choose to use a Hough transform rather than RANSAC to recognize clusters of 3 consistent features? (Note that RANSAC is more efficient in comparison with Hough) Link ...
0
votes
1answer
223 views

Combining custom YOLO network for face detection with another CNN

I am looking for a way to build and train an end-to-end CNN that contains two steps: 1) a CNN for finding a face and hands in the image and 2) CNN that works on the crops of the face and hands. To ...
0
votes
0answers
15 views

Is there any papers about learned image features in an unsupervised way?

I know that image features produced by pretrained models like VGG-net are the common way to initialize other networks in other tasks. But I was wondering, Is there image features that have been ...
0
votes
0answers
16 views

Different random weight initilization leading to different performances

I'm training a 3D U-Net on an EM dataset of a brain. The objective is to segment neurons in it. During the experiments, I've noticed, different random initialization of the network leads to different ...
0
votes
2answers
441 views

How to train CNN to detect single object

I want to train CNN using keras to detect a single object - "cars". Do I need to train CNN using cars' images only or need to train it using negative images (no cars) as well?
0
votes
0answers
19 views

Case where transfert learning performs better than finetuning

I'm working on a computer vision classification subject using CNN. I use a pretrained model as basis of my model. I read in Stanford CS 231n class that for large dataset with domain specific, best ...
0
votes
1answer
92 views

Advice needed for generating synthetic valence-arousal annotations?

For developing models on better understanding of visual image advertisements on an emotion perspective, my current dataset is having multi-labelled adjectives. Using an emotion dictionary which ...
0
votes
1answer
21 views

What technique can I best use to monitor system health status of camera system?

Introduction: I have a camera system (6 cameras) on basis of which a third party system performs classification and segmentation of images. This is used to monitor product quality of a products with ...
0
votes
1answer
108 views

Best ML technique for detecting multiple game cards in image

This question is related to a new hobby project i want to start. I have some experience with ML techniques and neural networks, although only for regression problems as of now. In my classification ...
0
votes
0answers
99 views

Classification of cereal boxes

I have a database of labeled images of cereal boxes and am doing some object recognition, intending to do deep learning. All the boxes have similar square shapes, but have some different colors, ...
0
votes
0answers
480 views

PCA on integers or floating numbers: is there any difference?

I want to use PCA to reduce SIFT (128 dims) and VLAD vectors (128*k dims, let's say ...
0
votes
0answers
642 views

imageNet: ground truth in classification task

I hope someone working with the www.image-net.org dataset reads this question. I am confused on what the ground truth is for the image classification task and how exactly the predictions of an ...
0
votes
1answer
94 views

SURF algorithm failing?

I am currently working on misalignment correction and after some research found a nice matlab toolbox for this purpose capable of doing Lucas-Kanade, ECC among others. I had success when using a ...

1
4 5 6
7
8