Ron Mokady (@mokadyron) 's Twitter Profile
Ron Mokady

@mokadyron

Visual Generative Models
Research lead @ BRIA AI

ID: 1247519499852468224

linkhttp://rmokady.github.io calendar_today07-04-2020 13:40:14

419 Tweet

1,1K Takipçi

355 Takip Edilen

fal (@fal) 's Twitter Profile Photo

Bria’s Image BG Removal was just the beginning - now Bria 3.2 and Video Background Removal are now on fal. Let’s create!

Bria’s Image BG Removal was just the beginning - now Bria 3.2 and Video Background Removal are now on fal. Let’s create!
Ron Mokady (@mokadyron) 's Twitter Profile Photo

Very excited about our ICCV work 🏄‍♂️ My favorite part: instead of using inversion, we just train the model to perform the identity function and used the emerged internal features at inference. Check the thread below for more details 😀

Misha Feinstein (@mishafein) 's Twitter Profile Photo

Classic background removal uses binary masks, and that’s the problem. Real-world edges aren't just black or white. Soft alpha matting assigns per-pixel transparency and changes everything. Read my latest blog to see how we built it BRIA AI: go.bria.ai/45gyW7L

apolinario 🌐 (@multimodalart) 's Twitter Profile Photo

ok i can't take it anymore: announcing the chatgpt image yellow tint corrector a Hugging Face space that runs locally on your browser to fix the yellow tint of the chatgpt generated images

ok i can't take it anymore: announcing the chatgpt image yellow tint corrector

a <a href="/huggingface/">Hugging Face</a> space that runs locally on your browser to fix the yellow tint of the chatgpt generated images
Ron Mokady (@mokadyron) 's Twitter Profile Photo

From my meetup talk a month ago: The biggest problem in text-to-image isn’t scaling or architectures. It’s evaluation. Today, we evaluate with “arenas”: user preference on short prompts. But humans struggle to follow long instructions, so to win you just optimize for average

sway (@swaystar123) 's Twitter Profile Photo

In the "Align your Flow" distillation paper, some of their example samples suffer from being overly/unnecessarily detailed. You can see this most prominently in their Appendix H.1. I believe this is because they distill at resolution 512 and then inference at a higher resolution

In the "Align your Flow" distillation paper, some of their example samples suffer from being overly/unnecessarily detailed. You can see this most prominently in their Appendix H.1.

I believe this is because they distill at resolution 512 and then inference at a higher resolution
BRIA AI (@bria_ai_) 's Twitter Profile Photo

Bria’s enterprise-grade Visual GenAI 3.2 Text-to-Image model is now natively integrated with 🤗 Diffusers 🧨, no custom pipeline required. 💻 Built on 100% licensed data ⚡️ Fast Implementation 🔒Commercial & IP-safe Explore now: 🤗 Hugging Face — Model weights, docs,

Bria’s enterprise-grade Visual GenAI 3.2 Text-to-Image model is now natively integrated with 🤗 Diffusers 🧨, no custom pipeline required.

💻 Built on 100% licensed data
⚡️ Fast Implementation
🔒Commercial &amp; IP-safe

Explore now:
🤗 Hugging Face — Model weights, docs,