CHIP: Adaptive Compliance for Humanoid Control through Hindsight Perturbation

์ €์ž: Sirui Chen, Zi-ang Cao, Zhengyi Luo, Fernando Castaรฑeda, Chenran Li, Tingwu Wang, Ye Yuan, Linxi "Jim" Fan, C. Karen Liu, Yuke Zhu | ๋‚ ์งœ: 2026-02-09 | DOI: 10.48550/arXiv.2512.14689 📄 PDF


Essence

Figure 1

Fig. 1: CHIP enables humanoid robots to perform manipulation tasks that require force control, such as wiping a whiteboa

CHIP๋Š” hindsight perturbation์„ ํ†ตํ•ด humanoid robot์ด ๋ฏผ์ฒฉํ•œ ์›€์ง์ž„์„ ์œ ์ง€ํ•˜๋ฉด์„œ๋„ ์ ์‘์  compliance๋ฅผ ๊ฐ–์ถ˜ forceful manipulation์„ ์ˆ˜ํ–‰ํ•  ์ˆ˜ ์žˆ๊ฒŒ ํ•˜๋Š” plug-and-play ๋ชจ๋“ˆ์ด๋‹ค.

Motivation

Achievement

Figure 1

Fig. 1: CHIP enables humanoid robots to perform manipulation tasks that require force control, such as wiping a whiteboa

How

Figure 2

Fig. 2: Overview of training and deployment of CHIP: this general formulation takes tracking targets, end-effector compl

Originality

Limitation & Further Study

Evaluation

Novelty: 4/5 Technical Soundness: 3/5 Significance: 4/5 Clarity: 4/5 Overall: 4/5

์ดํ‰: CHIP๋Š” humanoid์˜ agile motion๊ณผ compliant manipulation์„ ์–‘๋ฆฝ์‹œํ‚ค๋Š” ์šฐ์•„ํ•œ ํ•ด๊ฒฐ์ฑ…์œผ๋กœ, hindsight perturbation์ด๋ผ๋Š” ํ•ต์‹ฌ ์•„์ด๋””์–ด์˜ ๋‹จ์ˆœํ•จ๊ณผ ๊ธฐ์กด framework์™€์˜ ํ˜ธํ™˜์„ฑ์ด ๊ฐ•์ ์ด๋‹ค. ๋‹ค๋งŒ ์‹ค์ œ ๋กœ๋ด‡ ๊ฒ€์ฆ๊ณผ force control์˜ ์ •๋Ÿ‰์  ๋ถ„์„์ด ๋ณด์™„๋˜๋ฉด ๋”์šฑ ์™„์„ฑ๋„ ์žˆ๋Š” ์—ฐ๊ตฌ๊ฐ€ ๋  ๊ฒƒ์ด๋‹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˜์ •