G-NeuroDAVIS: A Neural Network model for generalized embedding, data visualization and sample generation

TL;DR

G-NeuroDAVIS has outperformed the Variational Autoencoder significantly in multiple key aspects, including embedding quality, classification performance, and sample generation capability, highlighting the potential of the generative model to serve as a powerful tool in various applications requiring high-quality data generation and representation learning.

Abstract

Visualizing high-dimensional datasets through a generalized embedding has been a challenge for a long time. Several methods have shown up for the same, but still, they have not been able to generate a generalized embedding, which not only can reveal the hidden patterns present in the data but also generate realistic high-dimensional samples from it. Motivated by this aspect, in this study, a novel generative model, called G-NeuroDAVIS, has been developed, which is capable of visualizing high-dimensional data through a generalized embedding, and thereby generating new samples. The model leverages advanced generative techniques to produce high-quality embedding that captures the underlying structure of the data more effectively than existing methods. G-NeuroDAVIS can be trained in both supervised and unsupervised settings. We rigorously evaluated our model through a series of experiments, demonstrating superior performance in classification tasks, which highlights the robustness of the learned representations. Furthermore, the conditional sample generation capability of the model has been described through qualitative assessments, revealing a marked improvement in generating realistic and diverse samples. G-NeuroDAVIS has outperformed the Variational Autoencoder (VAE) significantly in multiple key aspects, including embedding quality, classification performance, and sample generation capability. These results underscore the potential of our generative model to serve as a powerful tool in various applications requiring high-quality data generation and representation learning.

Figures (8)

• Figure 1: The network architectures of G-NeuroDAVIS for visualization of high-dimensional datasets.
• Figure 2: 2-dimensional embeddings produced by G-NeuroDAVIS (left) and VAE (right) for (A) $MNIST$ and (B) $FMNIST$ datasets respectively, in the context of unsupervised learning (Different class samples are represented by different colors).
• Figure 3: 2-dimensional embeddings produced by G-NeuroDAVIS (left) and VAE (right) for (A) $Coil20$ and (B) $OlivettiFaces$ datasets respectively, in the context of unsupervised learning (Different class samples are represented by different colors).
• Figure 4: Classification performance of G-NeuroDAVIS against VAE on (A) $MNIST$ and (B) $FMNIST$ dataset using k-NN and RF classifiers in terms of Accuracy, precision, Recall and F1-Score. Results have been obtained by varying parameter of k-NN ($neighbours$) and RF ($estimators$ / $trees$) respectively.
• Figure 5: Classification performance of G-NeuroDAVIS against VAE on (A) $Coil20$ and (B) $OlivettiFaces$ dataset using k-NN and RF classifiers in terms of Accuracy, precision, Recall and F1-Score. Results have been obtained by varying parameter of k-NN ($neighbours$) and RF ($estimators$ / $trees$) respectively.