AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

RealAppiance: Let High-fidelity Appliance Assets Controllable and Workable as Aligned Real Manauls

Yuzheng Gao, Yuxing Long, Lei Kang, Yuchong Guo, Ziyan Yu, Shangqing Mao, Jiyao Zhang, Ruihai Wu, Dongjiang Li, Hui Shen, Hao Dong

CVPR 2026 · Vision-language and multimodal › Robot action and navigation · approach: Memory and subgoal-based manipulation

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
proprietary MLLMs vs open-source MLLMsMLLMs · RealAppliance-Bench Task 1F1 (%): GPT-5 80.89 vs Qwen3-VL 235B Thinking 80.06; GLM-4.5V 73.94Improved
use embodied planning model (RoboBrain 2.0-7B) instead of general MLLMMLLM · RealAppliance-Bench Task 1Recall (%): 86.50 -> 5.78Hurt
use Gemini/GPT proprietary MLLMs for open-loop planningMLLMs · RealAppliance-Bench Task 2task success rate (%): GPT-5 1.22; Gemini 2.5 Flash 2.65No clear effect
use general MLLM for part groundingGPT-5 · RealAppliance-Bench Task 3average IoU (%): 12.15 (mAP 8.59)No clear effect
full-process chained inferenceMLLMs · RealAppliance-Bench Task 5task success rate (%): GPT-5 0.00; ApBot 0.92Hurt
compare embodied RoboBrain 2.0-7B on closed-loop adjustmentRoboBrain 2.0-7B · RealAppliance-Bench Task 4step-wise success rate (%): 31.77 vs GPT-5 29.61No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy