All Questions

I made two different type models with tensorflow, one is Class type, the other is function type for learning Fashion MNIST. I think the structure of two models is exactly same, but the results are not....

I'm trying to fine tune a model for a segmentation task. To see if everything is working properly I'm trying to make my model overfit on a single image split in 16 different patches. So in my code, I ...

What is the appropriate way to implement a differentiable variant of L0 regularizer (count the non-zero values in a Conv layer / matrix) to keras layer? I was thinking of using r(x) = tanh(abs(f*x)) ...

I am working on a semantic segmentation problem where I have an input x ( CT image) to a deep learning model of shape (batch_size,1,256,256) and an output of shape (batch_size,2,256,256) where the ...

I am making or trying to make a model that can estimate intrinsic camera parameters. For this, I want to use a custom loss function that uses the estimated intrinsic camera parameters, and then ...

Overview My model is an encoder that has input Z and output x. I'm trying to use a total_loss that has both traditional supervised learning and regularization term(s). I have additional functions (...

I’m currently working on a lane detection project using instance segmentation and following this tutorial. I’m attempting to train an ENet model with a discriminative loss function, but I'm ...

I am tring to make and then save a model in python file, and then train the model in another file. This model uses a custom loss function that is in a seperate python file, so it can be imported into ...

I am having difficulty saving and reloading a neural network model when I use a custom loss function. For example, in the code below (which integrates the suggestions of the related questions here ...

I'm working on implementing a custom loss function in TensorFlow/Keras that incorporates both categorical cross-entropy loss and a penalty based on custom rules. These rules are designed to penalize ...

Problem I'm having troubles correctly adding physics-informed losses to my training code for my neural network. Background I have an encoder that takes an input curve, X(w), where w is an independent ...

I am trying to train an autoencoder using a custom hamming loss function. The code is as follows, import tensorflow as tf import matplotlib.pyplot as plt from tensorflow.keras.layers import Input, ...

I've trained a Keras model and obtained a graph with training and validation loss curves. The graph shows the following pattern: enter image description here Can someone help me interpret this graph? ...

I have the following problem: Given an image my model outputs a 3-dimensional binary vector that consists of three features that need to infer from the image. Since, I don't care about the first ...

From a simple vanilla GAN code I look from GitHub I saw this generator model with activation sigmoid: # Generator G = tf.keras.models.Sequential([ tf.keras.layers.Dense(28*28 // 2, input_shape = (...