- Pivot 5
- Posts
- MobileDiffusion: Rapid text-to-image generation on-device
MobileDiffusion: Rapid text-to-image generation on-device
1. MobileDiffusion: Rapid text-to-image generation on-device
MobileDiffusion is an efficient latent diffusion model designed for mobile devices, offering rapid text-to-image generation on-device. It uses DiffusionGAN for one-step sampling during inference, fine-tuning a pre-trained diffusion model while leveraging a GAN to model the denoising step. Tested on iOS and Android premium devices, MobileDiffusion can generate a 512x512 high-quality image in half a second.
The model's small model size of just 520M parameters makes it uniquely suited for mobile deployment. The research aims to overcome the challenges imposed by limited computational power of mobile devices by examining each constituent and computational operation within Stable Diffusion's UNet architecture.
Read the full story here
2. The Times Sues OpenAI and Microsoft Over A.I. Use of Copyrighted Work
The New York Times has sued OpenAI and Microsoft for copyright infringement, claiming that millions of articles from The Times were used to train chatbots that now compete with it. The lawsuit, filed in Federal District Court, claims that the defendants should be held responsible for billions of dollars in statutory and actual damages related to the unlawful copying and use of The Times's uniquely valuable works.
The lawsuit could test the emerging legal contours of generative A.I. technologies and carry major implications for the news industry, as many newspapers and magazines have been hobbled by readers' migration to the internet.
Read the full story here
3. Mastercard jumps into generative AI race with model it says can boost fraud detection by up to 300%
Mastercard has launched a new generative artificial intelligence model, Decision Intelligence Pro, to help banks assess suspicious transactions on its network. The model, powered by a proprietary recurrent neural network, can help financial institutions improve their fraud detection rates by up to 300% in some cases.
The model is trained on data from the roughly 125 billion transactions that go through the company's card network annually, helping the AI understand relationships between merchants and predict fraudulent transactions. Mastercard claims the new transaction decisioning technology can help financial institutions improve their fraud detection rates by 20%, on average.
Read the full story here
4. Apple's Vision Pro will have Microsoft Teams, Word, Excel and other 365 apps at launch
Apple's Vision Pro AR/VR headset will launch on February 2 with dedicated Microsoft 365 apps, including Teams, Word, Excel, PowerPoint, Outlook, OneNote, and Loop. The apps are built-in tools that take advantage of the headset, such as a custom immersive environment for PowerPoint and quick document movement using the headset's pinch and drag functionality.
Microsoft Word will feature a focus mode, while Teams and Zoom will use the headset's persona feature. Headset owners will also have access to Microsoft's AI-powered Copilot service, allowing natural conversation to instruct the digital assistant to create drafts, summarize documents, and generate PowerPoint presentations.
Read the full story here
5. TikTok loses Taylor Swift, Drake, and other major Universal Music artists
Universal Music Group has removed the music catalogs of artists it represents, including Taylor Swift, Drake, and Olivia Rodrigo, from TikTok after negotiations failed to extend its expired licensing agreement. The dispute centers on artist pay, generative AI, and online safety.
UMG accused TikTok of attempting to bully it into accepting a "bad deal" that didn't address concerns about adequate compensation for artists and songwriters, AI-generated music protections, and online safety. TikTok criticized UMG for putting its own greed above the interests of its artists and songwriters.
Read the full story here
Sponsored by TLDR
Want a byte-sized version of Hacker news?
Try TLDR's free daily newsletter. TLDR covers the most interesting tech, science, and coding news in just 5 minutes. No sports, politics, or weather.