AI Technology & Industry Review
In a new paper High-Fidelity Audio Compression with Improved RVQGAN, a Descript research team presents Improved RVQGAN, a high fidelity universal audio compression model that combines advances in high-fidelity audio generation and improved adversarial and reconstruction losses to achieve 90x compression of 44.1 KHz audio at only 8kbps bandwidth.
In the new paper Towards Realistic Generative 3D Face Models, a research team from Carnegie Mellon University and Meta proposes a 3D controllable generative model capable of generating high-resolution textures and capturing high-frequency details in facial geometry. Their proposed AlbedoGAN outperforms state-of-the-art baselines in facial shape reconstruction.
In the new paper 3D-FM GAN: Towards 3D-Controllable Face Manipulation, a team from Princeton University and Adobe Research presents 3D-FM GAN, a novel conditional GAN framework that enables precise 3D-controllable face manipulation with high photorealism and strong identity preservation without requiring any manual tuning or optimizations.
A NVIDIA and Aalto University research team presents StyleGAN3, a novel generative adversarial network (GAN) architecture where the exact sub-pixel position of each feature is exclusively inherited from the underlying coarse features, enabling a more natural transformation hierarchy and advancing GAN-based animation generation.
A research team from Leibniz University Hannover introduces World-GAN, a 3D generative adversarial network that aims to learn and generate structures directly in the 3D voxel space in Minecraft.
A research team from Adobe proposes Directional GAN (DGAN), a novel and simple approach for generating high-resolution images conditioned on expected semantic attributes, greatly simplifying the image content generating process for marketing campaigns, websites and banners.
This paper proposes a novel graph-constrained generative adversarial network, whose generator and discriminator are built upon relational architecture.
As Synced previously reported, these hyperrealistic images now flooding the Internet come from US chip giant NVIDIA’s StyleGAN, a generative adversarial network based face generator that performs so well that most people can’t distinguish its creations from photos of real people.
It’s known that deep neural network models — mostly discriminative models — are prone to making poor predictions with high confidence when they encounter data distribution scenarios that vary too much from their training data.
A Google Brain research group has published a paper proposing a new method for evaluating generative models by using insights from competitive games between human players.
PixelGAN is an autoencoder for which the generative path is a convolutional autoregressive neural network on pixels, conditioned on a latent code, and the recognition path uses a generative adversarial network (GAN) to impose a prior distribution on the latent code.
Minimization of the KL-divergence can be thought of as maximizing the likelihood ratio, which appears in numerous applications.
A tweet from Google DeepMind’s Nenad Tomasev, concerning a paper from the prestigious Montreal Institute for Learning Algorithms has sparked a well-known natural language processing expert Yoav Goldberg to tweet back: “…I really dislike this work.”
New autoencoder-like generative network, called Adversarial Generator-Encoder Networks (AGE Network), does not have any discriminators, which makes the entire architecture much simpler than some recently-proposed GANs, but with nearly the same-level performance
This paper proposes a general approach to achieve “Image-to-Image Translation” by using deep convolutional and conditional generative adversarial networks (GANs).
ArtGAN can generate images with abstract information like images with a certain art style.