463 questions
0
votes
2
answers
236
views
Rotation with Pytorch Torchvision but from two specific angles
I want to randomly rotate my images at 45 degrees or -45 degrees, but only these two, no in between. 50% chances for 45 and 50% for -45.
I know that I can use RandomRotation of torchvision, but if I ...
- 13
1
vote
0
answers
34
views
AugLy fails to augment a COCO annotated dataset
I have an image I want to augment. Its resolution is 8192x4230. I've labeled it with this tool and exported the annotations in COCO format. Now I need to pass this image and annotations to AugLy ...
- 19
0
votes
1
answer
514
views
Why is Albumentations Adding Excessive Noise to Augmented Images?
I am working on a pill detection project using YOLOv8 and applying Albumentations for data augmentation. However, some augmented images turn out with too much noise or distortion (example attached).
...
1
vote
0
answers
103
views
How can I apply data augmentation to MNIST images in R's keras package?
I'm fluent in R but a total beginner with deep learning. I'm trying to splice together two bits of code from the Chollet et al. Deep Learning with R book (2d ed), namely one for a CNN model for the ...
- 35
2
votes
1
answer
125
views
Does order of transforms applied for data augmentation matter in Torchvision transforms?
I have the following Custom dataset class for an image segmentation task.
class LoadDataset(Dataset):
def __init__(self, img_dir, mask_dir, apply_transforms = None):
self.img_dir = img_dir
...
- 33
0
votes
0
answers
56
views
the data augmentation doesn't work properly
according to me, the given code should create 5700 images, which is 10 times the number of original images, instead, when I check the shape of it, it only gives 1140, which is twice the number of ...
0
votes
2
answers
88
views
How to Apply Image Augmentations in TensorFlow Pipeline for Large Dataset?
I have a dataset of images, each containing a 1 to 5letter word. I want to use deep learning to classify the characters that make up the word in each image. The labels for these images are formatted ...
0
votes
1
answer
255
views
Lack of libraries in textattack.augmentation
I am trying to augment my text data using textattack.augmentation (running on Kaggle).
The error is ImportError: cannot import name 'IdentityCache' from 'botocore.utils' (/opt/conda/lib/python3.10/...
1
vote
0
answers
27
views
Can I train a seq2seq model in which the same input sequence might have many possible output?
I'm trying to figure out a way to augment data for Seq2Seq model. But i have a limit in number of training samples. I have 200 samples.
I have an idea that using chat GPT to generates similar output ...
- 41
1
vote
1
answer
247
views
Albumentations intensity augmentations disrupt the image
I'm using a preprocessed, z-score normalized list as the source for my dataset.
Here's a collage of images augmented by Albumentations:
enter image description here
Here's my Compose:
augmentation = A....
- 13
-1
votes
1
answer
120
views
Resample filter in weka
I have a low amount of data instances in my data set. So, I tried the "resample" filter in Weka to increase the data amount and thus enhance the model performance. Is it okay to set the ...
- 1
0
votes
1
answer
360
views
How to use PyTorch transform to apply the same transform on the input/output image pairs?
I want to build a deep learning model by PyTorch that makes some image enhancements. Input and output of the model are the same size images.
I want to use PyTorch's torchvision.transforms library for ...
- 287
-4
votes
1
answer
156
views
Give more weights to some input features for variational auto encoder
I made a variation auto encoder to augment data.
It is working fine, but I would like some columns in the augmented data to have more variations than the others.
For example, I actually have this ...
- 2,074
-1
votes
1
answer
227
views
Should Augmented Data be Added to Existing Data or Used as Complete Replacement in CNN Models?
I need help with the optimal approach for integrating augmented data into Convolutional Neural Network (CNN) models. Specifically, should augmented data be added to the existing dataset to expand its ...
0
votes
0
answers
266
views
Randomness in Keras/Tensorflow RandomFlip (data augmentation)
In Keras/Tensorflow, the RandomFlip layer used in data augmentation, it us supposed that the operation is applied randomly to each image as it passes through the layer during training.
https://devdocs....
- 109