Stack Overflow
Stack Overflow

Questions tagged [image-classification]

579 questions
0
votes
1 answer

How to use Transfer learning to a non-related problem

I would like to know and understand if it's possible to use transfer learning to recognize images not related to the original model. For example, if I would like to recognize skin disease can I use a pre-trained model using a model that classifies…
bennius
  • 111
  • 7
0
votes
1 answer

Can I use string values as my dependent variable in KNN machine learning model?

So, I have the data with 128 face encodings with label being the name of the person and the column with names on it is my target column. I obviously used labelbinarisation to binarise the dependent variable(name column). When I used KNN to fit and…
BrightMango
  • 48
  • 9
0
votes
0 answers

How to create a multi dimensional NumPy Array for multi classes in deep learning?

I am following a siamese network example for face recognitions and it uses only two classes, with 3 images per each class (i.e. 6 total images). How can i modify following code to read and label into 6 image classes, with 10 images per each class…
Lakwin Chandula
  • 159
  • 1
  • 3
  • 15
0
votes
1 answer

Keras model performs well in training, validation and also in evaluating but performs poor when predicting a new image

I have trained a image classification model using keras which gives training acc of 98%, validation acc of 98% testing acc of 90%, but performs very poor on new input images. I don't know the why??
Bilal Ahmed
  • 17
  • 2
0
votes
0 answers

how to create our own dataset for tensorflow?

I'm new at Tensorflow. I have large image dataset. And I'm trying to create dataset for Tensorflow. I have searched a lot of resources but somehow couldn't implement on my dataset. Is there anyone who can help?
ecusbucus
  • 25
  • 1
  • 7
0
votes
0 answers

is there a non AI way of classifying images into two groups in Python?

I'm in the process of developing a parking app and need a way of classifying images of parking signs in the UK into their two colour types, yellow and white. I've tried a number of methods in openCv and the reliability that I've found just isn't…
DCharlie
  • 11
0
votes
2 answers

Firebase ML image classification probabilities got a weird output (not in probabilities) ----Update 1

I'm new to ML kit Firebase, I'm using the Firebase image classification custom model ML kits. There have 6 different objects in my custom model The custom model .tflite file was tested in python and it works perfectly. print(np.min(first_image),…
WHYLikeThis
  • 15
  • 5
-1
votes
0 answers

How to apply shap Explainable AI on my custom dataset

I am trying to apply shap Explainable AI on my custom dataset while using the pytorch in my python code. I found this tutorial on the shap…
Sanaa Ibraham
  • 1
-1
votes
0 answers

Trying to Find the Path for the Training Data Folder for an Image Classification Model in Python

I hope that you are doing well. I am trying to find the path to my training data file to build an Image Classification Model in Python with Jupyter Notebook.However, Python is giving me an Errno 13: Permission Denied:…
DBJ Robinson
  • 19
  • 2
-1
votes
0 answers

choosing the parameters for final model after crossvaligdation

I am training a tensorflow classification model with 3 fold cross-validation. Each fold reached max val_auc at different epochs. Now I want to use the entire dataset to train the final model. I am wondering what epoch should I choose?
user12967451
  • 29
  • 2
-1
votes
1 answer

How to handle Data Augmentation?

I'm building a CNN to perform a binary image classification. However, my dataset is very unbalanced with 10000 images from one class and the oter contains only 3000 (more or less). That's how my model is looking so far: model_1 = Sequential([ …
Vinicius Cavalcante
  • 1
  • 1
-1
votes
1 answer

CNN to classify images from a small dataset

I am trying to build a model that can classify a set of images. I think my main problem is that I have few images in the test set: 100 images. For the training phase I kept a percentage of 80% for training (310 images) and 20% for validation (78…
bscalingi
  • 1
  • 1
-1
votes
1 answer

Tensorflow 2.0 model always predicting the same label

I am pretty new to ML and tensorflow, so I was working through the tutorials on tensorflow and copy/pasted one of their models with a custom training loop with a few modifications. Upon training it on the cifar10 dataset over 3 epochs, it achieved a…
girthquake
  • 7
  • 3
-1
votes
1 answer

Issue with Tensorflow: model not learning anything, constant loss, unstable accuracy, validation accuracy exact 0

I am trying to train a small network for detecting JPEG artefacts, I have 7 classes with each 1k of the same images just with different level of artefacts. My code looks like this: import os import matplotlib.pyplot as plt import numpy as np import…
Josef Jílek
  • 3
  • 2
-1
votes
1 answer

Detect dark vs well-lit image using existing AI model

I am using the below code to detect dark vs well-lit images but it is not giving accurate results and is giving very long floating values like .00000000123 etc. #!/usr/bin/env python3 import cv2 import tensorflow as tf import numpy as np import…
Ayur Sharma
  • 31
  • 4
1 2 3
…
38 39