They ablate the embeddings in the paper and found that T5 and CLIP together are superior to either on their own. Not clear just how much difference it makes on text though… Imagen is bigger than SD, hard to disentangle the size from embedding on text. Good hypothesis though!
MULTIMODAL AI
-
Google Releases Universal Video Quality Model Open-Source
By
–
Today we announce the open-source release of the Universal Quality (UVQ) model. Read the blog to learn more, and be sure to check out the model → https://
github.com/google/uvq -
DAAM: Interpreting Stable Diffusion Cross Attention
By
–
What the DAAM: Interpreting Stable Diffusion Using Cross Attention Tang et al.: https://
arxiv.org/abs/2210.04885 #ArtificialIntelligence #DeepLearning #MachineLearning -
AI Dominates TIME’s Best Inventions List for 2022
By
–
0% surprised that @OpenAI Dall-E 2 and @NVIDIA Instant NeRF made it onto the Best Inventions of 2022 list from @TIME
. Artificial intelligence has been its own category on the list since 2020. Expect it to grow. -
Diffusion Models for Audio: Denoising as Subtractive Synthesis
By
–
Always thought diffusion was best suited for audio. Denoising = subtractive synthesis.
-
Metaverse Empathy Training Research Shows Promise
By
–
Yes I think shaping in areas like use of the metaverse for empathy training etc is especially interesting, really encouraging research in this area. Great share
-
SegCLR: Contrastive Learning for Cellular Morphology Analysis
By
–
Announcing Segmentation-Guided Contrastive Learning of Representations (SegCLR), a method that trains rich, generic representations of cellular morphology and ultrastructure without manual effort and can identify cell types from only small cell fragments→ https://t.co/7Ol91ZAhdx pic.twitter.com/Ff7qCYczxp
— Google AI (@GoogleAI) 9 novembre 2022Announcing Segmentation-Guided Contrastive Learning of Representations (SegCLR), a method that trains rich, generic representations of cellular morphology and ultrastructure without manual effort and can identify cell types from only small cell fragments→ https://
goo.gle/3G2uIEh -
Databricks AI-Powered Holiday Recipe Finder App
By
–
Did someone say…pie? Meet the Databricks Holiday Recipe Finder, an #AI-powered app that leverages text and image-based search to help you easily find the best recipes to add to your holiday table! What new recipes will be trying? https://
dbricks.co/3Ej5foT -
Photo Appears Similar to AI Generated Images
By
–
Apologies if I’m wrong, but the photo looks similar to AI type photos.
-
Bernard 2.0 interviews Polar, digital popstar avatar
By
–
Bernard 2.0 spoke with Polar, a digital popstar 🎶
— Bernard Marr (@BernardMarr) 9 novembre 2022
Did you miss the live broadcast?
If so, you can learn more about her creators, #TheSoulPublishing, and watch the full interview by clicking here 🤩 > https://t.co/zWRvcsQwuY#DigitalAvatar #VirtualInfluencer #Metaverse pic.twitter.com/Yeztve8uFXBernard 2.0 spoke with Polar, a digital popstar Did you miss the live broadcast? If so, you can learn more about her creators, #TheSoulPublishing, and watch the full interview by clicking here > https://
bit.ly/3WiCyzn #DigitalAvatar #VirtualInfluencer #Metaverse