Applyed regularization techniques and improvised the performance of VAE Model. such as L1/L2 Regularization (Weight Decay), Dropout, Batch Normalization, Beta-VAE (Modified KL Divergence Term), Data Augmentation Variable Auto Encoders: https://www.geeksforgeeks.org/variational-autoencoders/
Autoencoders are neural network architectures that are intended for the compression and reconstruction of data. It consists of an encoder and a decoder; these networks are learning a simple representation of the input data. Reconstruction loss ensures a close match of output with input, which is the basis for understanding more advanced architectures such as VAEs.
Dataset: CIFAR 10 Dataset