We prove identifiability of a broad class of deep latent variable models that (a) have universal approximation capabilities and (b) are the decoders of variational autoencoders that are commonly used in practice. Unlike existing work, our analysis does not require weak supervision, auxiliary information, or conditioning in the latent space. Specifically, we show that for a broad class of generative (i.e. unsupervised) models with universal approximation capabilities, the side information $u$ is not necessary: We prove identifiability of the entire generative model where we do not observe $u$ and only observe the data $x$. The models we consider match autoencoder architectures used in practice that leverage mixture priors in the latent space and ReLU/leaky-ReLU activations in the encoder, such as VaDE and MFC-VAE. Our main result is an identifiability hierarchy that significantly generalizes previous work and exposes how different assumptions lead to different "strengths" of identifiability, and includes certain "vanilla" VAEs with isotropic Gaussian priors as a special case. For example, our weakest result establishes (unsupervised) identifiability up to an affine transformation, and thus partially resolves an open problem regarding model identifiability raised in prior work. These theoretical results are augmented with experiments on both simulated and real data.

Did you miss a session from MetaBeat 2022? Head over to the on-demand library for all of our featured sessions here. Text-to-image AI systems such as DALL-E 2, Imagen and Midjourney are growing in popularity and capability right now, offering creators a revolutionary new way to produce content. Generating images from text prompts is a radical new approach to art-making and creative expression. But it also gives us the first glimpse of a fundamental shift in how we can better communicate and collaborate with our machines.

This article is part of a VB special issue. Read the full series here: How Data Privacy Is Transforming Marketing. We're putting another year of exciting developments in artificial intelligence (AI) deep learning behind us – one filled with remarkable progress, controversies and, of course, disputes. As we wrap up 2022 and prepare to embrace what 2023 has in store, here are some of the most notable overarching trends that marked this year in deep learning. One theme that has remained constant in deep learning over the past few years is the drive to create bigger neural networks.

Microsoft Bing is getting an AI image generator in the coming weeks, which allows users to turn text into digital art. Let's say a picture of a Shiba Inu as an astronaut would go perfectly with a blog post you're writing. You turn to the search engines for a free-to-use image, but you can't find one that matches your criteria. With the new Image Creator tool coming to Microsoft Bing, you can generate the exact image you need by inputting descriptive text. Image Creator is powered by DALL-E 2 image generator technology developed by OpenAI.

Developing deep generative models has been an emerging field due to the ability to model and generate complex data for various purposes, such as image synthesis and molecular design. However, the advancement of deep generative models is limited by challenges to generate objects that possess multiple desired properties: 1) the existence of complex correlation among real-world properties is common but hard to identify; 2) controlling individual property enforces an implicit partially control of its correlated properties, which is difficult to model; 3) controlling multiple properties under various manners simultaneously is hard and under-explored. We address these challenges by proposing a novel deep generative framework, CorrVAE, that recovers semantics and the correlation of properties through disentangled latent vectors. The correlation is handled via an explainable mask pooling layer, and properties are precisely retained by generated objects via the mutual dependence between latent vectors and properties. Our generative model preserves properties of interest while handling correlation and conflicts of properties under a multi-objective optimization framework. The experiments demonstrate our model's superior performance in generating data with desired properties.

Did you miss a session from MetaBeat 2022? Head over to the on-demand library for all of our featured sessions here. Today Microsoft Azure announced a variety of enhancements across its AI services at Ignite 2022. The company says the updates will help people "work smarter, not harder by bringing more intelligence, insights and value to the hands of customers." The product updates include new innovations in Azure Applied AI Services to help customers automate mundane tasks and serve end-users in multiple languages worldwide; updates to Azure Cognitive Services to "enrich and simplify" the creation of AI apps with pre-built models and text-to-image generation; and new capabilities in Azure Machine Learning that boost the productivity of developers and data scientists of all skill levels, and help further responsible AI deployment.

If you've ever been around kids, you know how limitless their imaginations can be. A simple cardboard box is an endless world of possibilities. What if we let that imagination run wild with DALL-E and Stable Diffusion? That's the amazing thing about the many AI image generators that are available now. Obviously, there are limits to what the AI can do, but you are basically free to enter anything and see what happens.

Everyone experiences grief at some point in their lives, whether it's when a relative, friend, or pet passes away.… Many often find comfort in keeping their memories of a loved one alive in some way. As technology progresses, a few have found solace in using artificial intelligence to reconnect with the dead. Generative AI offers imaginative ways to remember people's lives by simulating their likeness. The story of how one man primed a GPT-3-powered chatbot with text messages from his dead fiancée so that he could talk to her again went viral last year.

OpenAI's recent release of Whisper boasts human-level robustness and accuracy in speech recognition. I'm not Scottish (although I was born pretty close), but I immediately wanted to test it with a Scottish accent and compare it to "human-level". Having bought an unexciting new iPhone, at least I could put its A16 Bionic chip with 16-core Neural Engine through its paces for my experiment. Once the boring tech stuff was out of the way, I shared the test app on TestFlight with a few colleagues, yielding much amusement with its borderline magical results. Here's a little clip from the start of Trainspotting, which is particularly challenging for machines to understand; a Scottish accent over the top of Iggy Pop isn't something you'd train for.

It wasn't meant to happen like this. Yes, the robots were always going to come for everyone's jobs, but it was the menial ones that were set to go first. Freed from the need to fill out spreadsheets and perform administrative duties, we were all supposed to have extra time to indulge in more creative, fulfilling pursuits. Yet Microsoft Excel still exists while AI algorithms are producing works of art that are both commercially viable and critically respected. An AI artist, Jason Allen, recently caused outrage among old-school digital artists by winning a digital art competition. One of the writers of US publication The Atlantic, Charlie Warzel, provoked the ire of illustrators around the world by choosing to adorn an article about controversial radio host Alex Jones with an AI-generated caricature as opposed to using a stock photo or commissioning a portrait.