ZJU-OmniAI

non-profit

AI & ML interests

None defined yet.

Recent Activity

zwq2018 submitted a paper 4 days ago

VLA-Rolling: Fewer Denoising Steps, Better Control via Adaptive Action Reuse

zwq2018 submitted a paper 5 days ago

Scene-Designer: Learning Design Intelligence for Agentic 3D Scene Generation

zwq2018 submitted a paper 8 days ago

VLA-Δ: Learning Physical State Transition for Vision-Language-Action Models

View all activity

Papers

VLA-Rolling: Fewer Denoising Steps, Better Control via Adaptive Action Reuse

Scene-Designer: Learning Design Intelligence for Agentic 3D Scene Generation

View all Papers

OmniAI-ZJU 's papers 9

Submitted by

Wenqi Zhang

VLA-Rolling: Fewer Denoising Steps, Better Control via Adaptive Action Reuse

OmniAI-ZJU

Submitted by

Wenqi Zhang

Scene-Designer: Learning Design Intelligence for Agentic 3D Scene Generation

OmniAI-ZJU

Submitted by

Wenqi Zhang

VLA-Δ: Learning Physical State Transition for Vision-Language-Action Models

OmniAI-ZJU

Submitted by

Lizhou Liang

EmbodiedMemory-Bench: Benchmarking Embodied Memory for Long-Horizon Embodied Tasks

OmniAI-ZJU

4

Submitted by

Wenqi Zhang

Spatial-Interactor: Learning Spatial Reasoning through Interaction with the Observable Physical World

OmniAI-ZJU

Submitted by

Wenqi Zhang

Embodied-Navigator: Point, Think, Memorize, and Align for Efficient Navigation

OmniAI-ZJU

Submitted by

Wenqi Zhang

Show, Don't Tell: Evaluating Spatial Cognition in Generative Pixels Rather Than LLM Text

OmniAI-ZJU

Submitted by

Wenqi Zhang

VLA-Corrector: Lightweight Detect-and-Correct Inference for Adaptive Action Horizon

OmniAI-ZJU

Submitted by

Wenqi Zhang

GFT: From Imitation to Reward Fine-Tuning with Unbiased Group Advantages and Dynamic Coefficient Rectification

OmniAI-ZJU