Today at 1:00 PM, visit the @ICCVConference Google booth to learn how research scientists at Google are using WHOOPS, a vision and language benchmark of synthetic and compositional images, to assess multimodal chatbots for visual commonsense.
MULTIMODAL AI
-
Google Preface: Ultra-High Resolution 4K Face Rendering from Few Images
By
–
Preface renders novel views of faces at ultra-high 4K resolution from very few input images. Stop by the #ICCV2023 Google booth today at 12:30 PM to learn how this approach generalizes to images captured in-the-wild using a mobile camera. Learn more →
-
Building Software with AI and GPT-4 Vision Learning Paths
By
–
Ok who’s building the course on how to learn to build software with AI, GPT4 vision etc Got to be better paths than 100 days of python etc now right?
-
Cloud-based Language Embedded Radiance Fields Demo at IROS23
By
–
Demo of cloud-based Language Embedded Radiance Fields (LERFs) at #iros23 in Detroit until 5pm today.
@BoschGlobal
-
Poe Launches Image Rendering API for Bot Image Generation
By
–
The other major improvement launching today is the ability to return images from a bot, and have Poe clients render them as first-class objects, as shown in SDXLx3 above, and in the StableDiffusionXL bot launched last week, which is also built entirely on this same API.
-
SDXLx3 Bot Generates Multiple Images via StableDiffusion
By
–
SDXLx3 is an example bot that will use StableDiffusionXL to generate 3 images at once for any given query (with no cost to the developer for the model inference). https://
poe.com/SDXLx3 -
Meta launches GenAI tools for marketers: image backgrounds and ad variations
By
–
Meta’s rolling out new #GenAI tools for marketers: A way to generate image backgrounds, vary text for ads and adjust image sizes for each format. Also worth mentioning other companies already offer similar GenAI tools, but could this help Meta cut out intermediaries?
-
Fast SDXL Inference with JAX on TPU v5e
By
–
Fast SDXL Inference with JAX on TPU v5e Amazing how fast anyone can generate 4 1024×1024 resolution SDXL images in under 5 seconds, for free: https://
huggingface.co/spaces/google/
sdxl
… -
In-Context Learning for Novel Subject Rendering Without Fine-Tuning
By
–
Drop by the #ICCV2023 Google booth today at 12:30 PM to learn about an in-context learning technique that, given examples of a new subject (e.g., sunglasses, a dog), generates novel renderings of it in different scenes, without the need for fine-tuning.
-
Quick Release Support for Audio Video Text and Multilingual
By
–
Cool to see this come out so quickly, and already supports audio, video, and text as well as multilingual. Looking forward to playing with it once it is released. Congrats @YiTayML
!