Size Wu

PhD student@NTU

[Size Wu @WuSize]

🔥 We release Harmon: a unified framework for multimodal understanding & generation with a shared visual encoder (vs. decoupled Janus/-Pro).

💥 SOTA on GenEval, MJHQ, WISE
🧠 Strong understanding performance
📄 Paper: https://t.co/RFhEl9NEN7
🔗 Code: https://t.co/O6og7NcYAI

[Size Wu @WuSize]

✅ Harmon models are fully converted to 🤗 HuggingFace format—easy to load and integrate into your pipelines. 👉https://t.co/FIPE93wgXK

[Size Wu @WuSize]

🔥 We release Harmon: a unified framework for multimodal understanding & generation with a shared visual encoder (vs. decoupled Janus/-Pro).

💥 SOTA on GenEval, MJHQ, WISE
🧠 Strong understanding performance
📄 Paper: https://t.co/RFhEl9NEN7
🔗 Code: https://t.co/O6og7NcYAI

[Size Wu @WuSize]

🖼️ Try the live demo here 👉 https://t.co/V7Cp9I9rUY
Supports multimodal inputs and text-to-image generation!

[Size Wu @WuSize]

@LucasJinReal @EIFY @arankomatsuzaki They used (shown in Figure3b). It seems they are distilling CLIP’ knowledge to the LLM.

[Size Wu @WuSize]

🌟Our paper proposes a clever and novel approach to Open-Vocabulary Object Detection. Come and check it in this afternoon #CVPR2023 ! ​

📍 Location: West Exhibit Hall #276
⏰ Time: 04: 30 PM - 06:00 PM

Paper: https://t.co/Pyki7tVThJ