Deep neural network models have witnessed significant success in modern computer vision pipelines. Deep Nostalgia utilises Generative Adversarial Networks (GANs) to animate faces in still photos. Here is a list of the top alternatives to Deep Nostalgia.
(The list is in no particular order)
1| Unfade
About: Unfade is an app that helps photographers immortalise their images. The app automatically analyses the photo and refreshes faded colours. According to sources, this app is particularly created to tune older photographs by detecting faded colours and applying intelligent filters automatically to coax them back to their former beauty. The app also crops photos and shares them straight to Apple Photos.
Know more here.
2| MoCoGAN
About: Motion and Content decomposed Generative Adversarial Network (MoCoGAN) is a framework for motion and content decomposed video generation based on Generative Adversarial Networks (GANs). The network generates videos from random inputs.
MoCoGAN adopts a motion and content decomposed representation for video generation. It uses an image latent space, where each latent code represents an image and divides the latent space into content and motion subspaces. For instance, given videos of people performing different facial expressions, MoCoGAN learns to separate a person’s identity from their expression, thus allowing us to synthesise a new video of a person performing different expressions or fixing the expression and generating various identities.
Know more here.
3| PhotoGlory
About: PhotoGlory is an old photo restoration software that uses artificial intelligence and neural networks to repair damaged photos and colorise black & white shots. According to sources, the AI-based software analyses the image content and detects faces, furniture, sky, trees etc. Then it comes up with a color palette best for this particular photo.
Know more here.
4| DeOldify
About: DeOldify is a deep learning-based project for colourising and restoring old images and videos. The Black and White image colourising library created by Jason Antic introduced the NoGAN technique to make hyper-realistic colourisation images and video. The features include video glitch elimination, less biasing for Blue color, more accurate skin tone and highly detailed and hyper-realistic outputs. DeOldify provides three primary models for different use cases: artistic model, stable model and video model.
Know more here.
5| Few-Show Adversarial Learning Framework
About: Few-shot learning is a popular technique in computer vision applications to classify data or images by using few to one example of the target subject. The approach used few shot learning to create talking head models from a handful of photographs and with limited training time. The framework is basically a meta-learning of adversarial generative models, which is able to train highly realistic virtual talking heads in the form of deep generator networks.
Know more here.
6| Deep Latent Space Translation
About: Deep Latent Space Translation is a method of restoring old photos. The deep learning approach comes with a new method called triplet domain translation network. In this technique, two variational autoencoders (VAEs) are trained to transform and clean old photos into two latent spaces. VAE1 is trained for images in real photos and synthetic images and VAE2 is trained for clean images.
Know more here.
