We present Corgi, a novel method for text-to-image generation. Corgi is based on our proposed shifted diffusion model, which achieves better image embedding generation from input text. Unlike the baseline diffusion model used in DALL-E 2, our method seamlessly encodes prior knowledge of the pre-trained CLIP model in its diffusion process by designing a new initialization distribution and a new transition step of the diffusion. Compared to the strong DALL-E 2 baseline, our method performs better in generating image embedding from the text in terms of both efficiency and effectiveness, resulting in better text-to-image generation. Extensive large-scale experiments are conducted and evaluated in terms of both quantitative measures and human evaluation, indicating a stronger generation ability of our method compared to existing ones. Furthermore, our model enables semi-supervised and language-free training for text-to-image generation, where only part or none of the images in the training dataset have an associated caption. Trained with only 1.7% of the images being captioned, our semi-supervised model obtains FID results comparable to DALL-E 2 on zero-shot text-to-image generation evaluated on MS-COCO. Corgi also achieves new state-of-the-art results across different datasets on downstream language-free text-to-image generation tasks, outperforming the previous method, Lafite, by a large margin.

Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. It's been an exciting few months since OpenAI released ChatGPT, which now has everyone talking about it, many talking to it and all eyes on what's next. ChatGPT raised the bar for what computers are capable of and is a window into what's possible with AI. And with tech giants Microsoft, Google and now Meta joining the race, we should all buckle up for an exciting but potentially bumpy ride. Core to these capabilities are large language models (LLMs) -- specifically, a particular generative LLM that makes ChatGPT possible.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. Adobe announced a new "family of creative generative [artificial intelligence] models" called Firefly this week. In a Monday release, the company said its first model would empower customers to generate high-quality images and text effects. "Adobe Stock's hundreds of millions of professional-grade, licensed images are among the highest quality in the market and help ensure Adobe Firefly won't generate content based on other people's or brands' IP," it said, also assuring that Adobe would continue to prioritize countering potential harmful bias as future Firefly models leverage a variety of assets, tech and data from Adobe and others.

Consumer demand for AI chatbot experiences has been funneling millions of dollars into mobile apps advertising their association with ChatGPT or OpenAI technologies. According to a new analysis of the AI app ecosystem from analytics provider data.ai, And that demand is continuing to grow. In February 2023, these 10 apps combined accounted for nearly $5.9 million in global consumer spending, the firm says. And within the first 20 days of March, the apps were averaging $232,000 in daily consumer spending, up 11% from the average of $210,000 in February.

Following on from the integration of ChatGPT into the Bing search engine, Microsoft have now followed up by integrating another of OpenAI's products: the AI image generator DALL-E 2. It's fair to say that rolling out the "new Bing" with its ChatGPT-powered AI chat functionality was a whopping success for Microsoft. Now, says Microsoft Corporate VP Yusuf Mehdi in a blog post (opens in new tab), the tech giant are "taking the chat experience to the next level by making the new Bing more visual." What that boils down to is utilizing the OpenAI's AI image generator, DALL-E 2, to form the Bing Image Creator. Essentially, instead of using DALL-E 2 to generate images on its own website, you can type prompts into the Bing search engine to receive AI generated imagery from there using the same engine. So how do you use the Bing Image Creator?

UNTIL Microsoft curtailed the capabilities of its Bing chatbot – codenamed Sydney and powered by an advanced version of OpenAI's ChatGPT model – there were a chaotic few days last month when it was threatening, cajoling, falling in love with and terrifying its beta testers. Even journalists who regularly write about artificial intelligence expressed surprise: they know these programs are just statistical models of the language on the internet, but they still found Sydney's "personality" unsettling and eerily human. Bing's chatbot has yet to be rolled out to the world at large and its curtailment has prevented it from going off the rails again, but it remains unnerving.

GTC--NVIDIA today announced an expanded set of generative AI cloud services for customizing AI foundation models to accelerate the creation of new proteins and therapeutics, as well as research in the fields of genomics, chemistry, biology and molecular dynamics. Part of NVIDIA AI Foundations, the new BioNeMo Cloud service offering -- for both AI model training and inference -- accelerates the most time-consuming and costly stages of drug discovery. It enables researchers to fine-tune generative AI applications on their own proprietary data, and to run AI model inference directly in a web browser or through new cloud application programming interfaces (APIs) that easily integrate into existing applications. "The transformative power of generative AI holds enormous promise for the life science and pharmaceutical industries," said Kimberly Powell, vice president of healthcare at NVIDIA. "NVIDIA's long collaboration with pioneers in the field has led to the development of BioNeMo Cloud Service, which is already serving as an AI drug discovery laboratory. It provides pretrained models and allows customization of models with proprietary data that serve every stage of the drug-discovery pipeline, helping researchers identify the right target, design molecules and proteins, and predict their interactions in the body to develop the best drug candidate."

After analyzing the images generated by DALL-E 2 and Stable Diffusion, they found that the models tended to produce images of people that look white and male, especially when asked to depict people in positions of authority. That was particularly true for DALL-E 2, which generated white men 97% of the time when given prompts like "CEO" or "director." That's because these models are trained on enormous amounts of data and images scraped from the internet, a process that not only reflects but further amplifies stereotypes around race and gender. But these tools mean people don't have to just believe what Hugging Face says: they can see the biases at work for themselves. For example, one tool allows you to explore the AI-generated images of different groups, such as Black women, to see how closely they statistically match Black women's representation in different professions.

Adobe is jumping into the generative AI game with the launch of a new family of AI models called Firefly. Focused on bringing AI into Adobe's suite of apps and services, specifically AI for generating media content, Firefly will be made up of multiple AI models "working across a variety of different use cases," Adobe VP of generative AI Alexandru Costin told TechCrunch in an email interview. It's an expansion of the generative AI tools Adobe introduced in Photoshop, Express and Lightroom during its annual Max conference last year, which let users create and edit objects, composites and effects by simply describing them. As the fervor around the tech grows, Adobe has raced to maintain pace, for example allowing contributors to sell AI-generated artwork in its content marketplace. "Firefly is the next step on our AI journey -- bringing together our new'gentech' models with decades of investment in imaging, typography, illustration and more to produce assets," Costin said.

Today, we are excited to announce that GPT-4 is available in preview in Azure OpenAI Service. Customers and partners already using Azure OpenAI Service can join the waitlist to access GPT-4 and start building with OpenAI’s most advanced model yet. With this milestone, we are proud to bring the world’s most advanced AI models—including GPT-3.5, ChatGPT, and DALL•E 2—to Azure customers, backed by Azure AI-optimized infrastructure, enterprise-readiness, compliance, data security, and privacy controls, along with many integrations with other Azure services.