Cognition to Control - Multi-Agent Learning for Human-Humanoid Collaborative Transport

์ €์ž: Hao Zhang, Ding Zhao, H. Eric Tseng | ๋‚ ์งœ: 2026-03-04 | URL: https://arxiv.org/abs/2603.03768 📄 PDF


Essence

Figure 3

Fig. 3: The proposed hierarchical HRC framework for humanoid-object coordination, partitioning decision-making into thre

์ธ๊ฐ„-ํœด๋จธ๋…ธ์ด๋“œ ํ˜‘์—… ์šด๋ฐ˜์„ ์œ„ํ•œ 3๊ณ„์ธต Cognition-to-Control ํ”„๋ ˆ์ž„์›Œํฌ๋กœ, VLM ๊ธฐ๋ฐ˜ ์˜๋ฏธ๋ก ์  ์ถ”๋ก , Markov potential game ๊ธฐ๋ฐ˜ MARL ์กฐ์ •, ์ „์‹  ์ œ์–ด๋ฅผ ํ†ตํ•ฉํ•˜์—ฌ ์—ญํ• ์˜ ์ž๋™ ํ˜•์„ฑ๊ณผ ๊ฐ•๊ฑดํ•œ ํ˜‘์—…์„ ์‹คํ˜„ํ•œ๋‹ค.

Motivation

Achievement

Figure 1

Fig. 1: Demonstration of human-robot collaboration via cognition-to-control hierarchy: (a) the humanoid and human partne

How

Figure 3

Fig. 3: The proposed hierarchical HRC framework for humanoid-object coordination, partitioning decision-making into thre

Originality

Limitation & Further Study

Evaluation

Novelty: 4/5 Technical Soundness: 4/5 Significance: 4/5 Clarity: 4/5 Overall: 4/5

์ดํ‰: ์ธ๊ฐ„-๋กœ๋ด‡ ํ˜‘์—…์˜ ๊ทผ๋ณธ์ ์ธ ์ธ์ง€-์ œ์–ด ๋‹จ์ ˆ ๋ฌธ์ œ๋ฅผ 3๊ณ„์ธต ๊ตฌ์กฐ๋กœ ์ฒด๊ณ„์ ์œผ๋กœ ํ•ด๊ฒฐํ•˜๊ณ , Markov potential game MARL์„ ํ†ตํ•ด ๋ช…์‹œ์  ์—ญํ•  ํ• ๋‹น ์—†์ด ํ˜‘์—… ์—ญํ• ์ด ์ž๋™ ํ˜•์„ฑ๋˜๋Š” novel ์ ‘๊ทผ๋ฒ•์„ ์ œ์‹œํ•œ๋‹ค. ์‹คํ—˜ ๊ฒฐ๊ณผ๋Š” ๊ฐ•๊ฑด์„ฑ๊ณผ ์œ ํšจ์„ฑ์„ ์ž˜ ๋ณด์—ฌ์ฃผ์ง€๋งŒ, ์ž‘์—… ๋‹ค์–‘์„ฑ ๋ฐ ํ™˜๊ฒฝ ์กฐ๊ฑด ๋ฒ”์œ„ ํ™•๋Œ€๊ฐ€ ํ•„์š”ํ•˜๋‹ค.

← ๋ชฉ๋ก์œผ๋กœ ๋Œ์•„๊ฐ€๊ธฐ

๐ŸŽง Audio Overview

์ด ๋…ผ๋ฌธ ๋ฆฌ๋ทฐ๋ฅผ ํŒŸ์บ์ŠคํŠธํ˜• ์˜ค๋””์˜ค๋กœ ์ƒ์„ฑํ•ฉ๋‹ˆ๋‹ค. (Gemini ยท ํ‚ค๋Š” ๋ธŒ๋ผ์šฐ์ €์—๋งŒ ์ €์žฅ ยท ์™„์„ฑ๋ณธ์€ ์ด๋ฉ”์ผ๋กœ๋„ ์ „์†ก)
โ–ธ ๊ณ ๊ธ‰: ๊ตฌ์„ฑ ๋ฐฉํ–ฅ(๋Œ€๋ณธ ์ž‘์„ฑ ์ง€์นจ) ์ง์ ‘ ์ˆ˜์ •