Stack Exchange Network

Stack Exchange network consists of 175 Q&A communities including Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.

Visit Stack Exchange

Questions tagged [computer-vision]

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

1
vote
0answers
9 views

Same input size for Source and Target Model?

For the transfer learning do we need to have same input image size of target model as source model? for example my source model is trained on 100x100 input images and target model is low resolution ...
0
votes
0answers
8 views

What types of problems can object detection not be used on?

I'm curious what types of problems can object detection(ex: Faster-RCNN) not be used? For example, I'm guessing if you were trying to detect an object which is conditioned on something else occurring ...
0
votes
0answers
10 views

YOLOv3 Detection Layer

I'm learning about YOLOv3 and wanted to confirm my understanding of how the detection layer works. Just to confirm, we have some output tensor from which we would filter for the objectness score above ...
0
votes
1answer
16 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
21 views

What is the best way to normalise image data?

The normalisation in an image really confuses me. I mean there are multiple ways to do it (see below) but, is there the best one, or most preferable one, or one needs to experiment with all to find ...
0
votes
0answers
28 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 (...
1
vote
2answers
42 views

In CNN, how to map from fully connected layer to output image?

In CNN, where the last layers are fully connected, how to make pixel-wise prediction to output an image(binary matrix), if the number of neurons in the last layer is less than the size of the image?
0
votes
0answers
9 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. ...
1
vote
0answers
32 views

What is the relation between a loss function and an energy function?

A loss function is a function that measures the distance between the expected value and the actual value of a model (an example of a loss function is the cross entropy). An energy function can be ...
0
votes
0answers
12 views

how to take a subset of a dataset to fine-tune a neural network?

I would like to build a classifier with 80 000 images and 45 classes. As each epochs takes 1 hour to train, Is there a way to win time by training only a subset of the dataset without lowering the ...
0
votes
0answers
22 views

How are YOLO anchor boxes generated?

I am recently trying out darkflow, a Tensorflow implementation of Darknet written by Joseph Redmon. Looking at the configuration files, I noticed a section called region as shown below. ...
0
votes
1answer
29 views

How best to combine object detection and tracking

I am trying to make a computer vision system which will be able to detect and track objects of interest. This will require (1) detection functionality to notice the object when it appears (2) ...
0
votes
0answers
17 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
1answer
43 views

Deep learning models for unsupervised semantic segmentation

I am working on semantic segmentation for satellite images using keras and python. It is my understanding that popular models like U-Net require mask images (labels). Are there any unsupervised deep ...
0
votes
0answers
15 views

Training a Semantic Segmentation Model with Partially Labeled Data

I have recently been tasked with a two-class semantic segmentation problem on aerial imagery. From what I can tell, off-the-shelf archetectures like U-Net seem to do well in this domain, so I plan on ...
1
vote
2answers
56 views

Pretext Task in Computer Vision

I am new to Computer Vision. I am reading many papers and i see the term "pretext task". Can anyone explain what exactly it means. Thanks in Advance.
0
votes
0answers
37 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
6 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 ...
3
votes
1answer
45 views

Are there any other image classification methods besides using neural networks?

When reading about image classification, the only occurring terms are "neural networks", "deep learning" and "CNN". It seems like there are no other methods for this task. I have worked with neural ...
0
votes
0answers
12 views

Normalising predictions across datasets

I am currently training a model to predict a binary attribute. The model gives the output in range [0, 1]. The metric is TPR@FPR, e.g. I need to achieve maximum ...
2
votes
1answer
12 views

Model to Recommend Ideal Parameter Changes for Best Performance of Industrial Machine

I'm trying to develop a machine learning model to solve this problem, and am unsure of where to start. We begin with some user-defined settings. The settings are used by a machine to create a product....
2
votes
0answers
10 views

What are the important methods that evolved in computing optical flow?

I have gone through various approaches to find optical flow. But I have a tad confusion between Horn and Shunck method and Lucas Kannede method. Where are these methods useful and where do these ...
1
vote
1answer
66 views

How can we relate the concepts of GAN/cGAN in SRGAN? Is SRGAN a Conditional GAN?

I have been reading and looking at implementations of the SRGAN, from "Photo-realistic Single Image Super Resolution with Generative Adversarial Networks" paper. One thing that I noticed is that the ...
0
votes
0answers
26 views

What is scale-invariance and log-space translations of a bounding box?

In slow R-CNN paper, the bounding box regression's goal is to learn a transformation that maps a proposed bounding box P to a ground-truth box G and we parameterize the transformation in terms of four ...
4
votes
1answer
242 views

Why do people use Zero-Padding in Convolutional Neural Networks?

I am wondering why people usually pad with zeros instead of e.g., using the min-value. Zero-padding, in my opinion, makes sense if you have input images with a pixel range [0, 255] or [0, 1] (after ...
1
vote
1answer
36 views

Simple but effective face detection algorithm using neural networks

I'd like to have my undergrad machine learning students have the option of doing a face detection project using neural networks (constructed by the students using Keras). The algorithm should ideally ...
0
votes
0answers
27 views

Optical Character Recognition - digits on the screen

My task is to classify a digit based on a small image containing one digit only. The font type and size is the same across the training/test dataset, but the position of the digit in the image might ...
0
votes
1answer
37 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 ...
2
votes
0answers
19 views

How can I lower the resources taken by an ANN while limiting data loss?

I designed a visual quality assurance system with a goal to determine whether a factory part during manufacturing is flawed or not. The need arose because it is rather difficult to tell if a part is ...
1
vote
1answer
21 views

Differentiable method to convert voxel representation into pointclouds?

I'm finding a way to convert 3D voxel data into 3D point-clouds. Since I'd like to do it inside a deep-learning architecture, the conversion has to be differentiable. Is there such a method? Thanks ...
-1
votes
1answer
90 views

this question is related to activation function

I could not understand : what is activation score from the kernel in the previous stage? I know what activation function mean ,activation functions type and how its work But what about activation ...
1
vote
0answers
16 views

Learning useful semantic representations of data

Training a neural network on its final task (e.g. classification) right from the beginning is not always the best way to go. I'd like to make a short list of recognized methods of motivating a NN to ...
1
vote
0answers
15 views

Stereo image classification [closed]

I just been wondering how can I combine stereo vision and Convolutional neural networks for a classification problem
0
votes
0answers
37 views

How do convolutional predictors work in SSD Object Detection?

I'm trying to understand this paper SSD: Single Shot MultiBox Detector by Liu et al, there they mention "Convolutional predictors for detection: Each added feature layer (or optionally an ex- ...
0
votes
0answers
25 views

How to reduce false positives when novel negative images are visually similar to training images?

I am noticing that my ResNet model is showing some false positives in cases where novel negative example images are somewhat visually similar to positive examples. In these cases, it's not simply ...
0
votes
0answers
20 views

Do convnets have issues detecting small features?

I remember seeing some presentation about how convnets had issues detecting say, glasses on a person's face because they take up very little pixel space. However no one seems to be saying that anymore....
0
votes
0answers
14 views

Can haar cascades work for hand recognition with multiple hand shapes?

I am researching about using haar cascades for hand recognition. There are examples on the internet (some from OpenCV) for detecting a fist or an open hand with haar cascades (at a time). My question ...
1
vote
0answers
23 views

Image generation based on sketch

Are there any instances of image generation models, where an image (a very rough sketch) has been used as an input and was then augmented. For example: This could be a rough sketch, which is then ...
8
votes
3answers
467 views

Is it possible to give variable sized images as input to convolutioal neural network

Can we give images with variable size as input to convolutional neural network for object detection? If possible, How can we do that? But if we try to crop the image, we will be loosing some portion ...
0
votes
0answers
19 views

Using the same image across multiple classes in image classification

I have a multi-label data set that I'm trying to use for multi-class image classification. Each image potentially has more than one class and is thus being selected as a positive example of as many ...
0
votes
1answer
72 views

When doing data augmentation, should you train with the original data as well or just the augmented data?

When doing data augmentation in computer vision problems, should you train with the original (un-augmented) data as well or just the augmented data? Are there pros and cons to the two strategies or ...
0
votes
1answer
24 views

How does smoothing an image gives it a different scale?

In some deep learning papers i read about multiscale inputs, so i wanted to read about scale of an image. What i got to know is that fundamentally scale is related to the distance of the object being ...
0
votes
0answers
20 views

Interpreting probabilities from image classifier, which model to use?

I'm trying to interpret examples from a probability perspective and my intuition is telling me Logistic Regression should be used for such a purpose despite the score being weaker than the other ...
0
votes
0answers
25 views

What is crop size for in semantic segmentation?

In Tensorflow Deeplabv3 I saw the training and validation parameter called crop_size, but they are different values in training and validation. If my network is trained on images of size 512x512, ...
1
vote
0answers
105 views

Translation invariance of features in convolutional encoder-decoders

A big part of using convolutional layers is translation invariance, i.e., features are detected regardless of their position in the image. In a convolutional encoder-decoder that maps from an image ...
0
votes
1answer
84 views

Introduction to Conditional random fields

I came across the application of a conditional random field (CRF) to the output from a convolutional neural network (CNN) for image segmentation. The additional CRF step seems to be a common ...
2
votes
1answer
112 views

Custom Loss Function - Inducing sparsity

From the comments, I realized that my question wasn't clear enough, so I'll start with a short background. I am trying to construct an attention model that performs classification based on just a ...
1
vote
0answers
549 views

What is the difference between dice loss vs jaccard loss in semantic segmentation task?

What is the difference between dice loss vs jaccard loss in semantic segmentation task? Dice loss: ...
0
votes
0answers
21 views

what will “Faster RCNN”(or any other object detection algorithm that uses anchors) do in this situation?

Can anyone please tell me what will "Faster RCNN"(or any other object detection algorithm that uses anchors) do in this situation? If there are 2 object and both are inside 2 different anchors and ...
1
vote
0answers
53 views

clustering objects in point cloud

I am currently working on point cloud data analysis, trying to label objects which are not ground or vegetation e.t.c. So far I tried many clustering algorithms, with moderate success. In my best ...