Diffusion Models for Robotic Manipulation: A Survey

์ €์ž: Rosa Wolf, Yitian Shi, Sheng Liu, Rania Rayyes | ๋‚ ์งœ: 2025-04-11 | URL: https://arxiv.org/abs/2504.08438 📄 PDF


Essence

Figure 1

Figure 1: Illustrations of diffusion (forward) processes on image, trajectories, and grasp poses (Urain et al. (2023)) a

๋ณธ ๋…ผ๋ฌธ์€ ๋กœ๋ด‡ ์กฐ์ž‘(robotic manipulation) ๋ถ„์•ผ์—์„œ diffusion model์˜ ์‘์šฉ์„ ์ข…ํ•ฉ์ ์œผ๋กœ ๋ฆฌ๋ทฐํ•˜๋Š” ์ฒซ ๋ฒˆ์งธ survey๋กœ, grasp learning, trajectory planning, data augmentation ๋“ฑ์˜ ์ฃผ์š” ์‘์šฉ ๋ถ„์•ผ์™€ ํ•™์Šต ํ”„๋ ˆ์ž„์›Œํฌ, ์•„ํ‚คํ…์ฒ˜๋ฅผ ์ฒด๊ณ„์ ์œผ๋กœ ๋ถ„๋ฅ˜ํ•œ๋‹ค.

Motivation

Achievement

Figure 1

Figure 1: Illustrations of diffusion (forward) processes on image, trajectories, and grasp poses (Urain et al. (2023)) a

How

Figure 1

Figure 1: Illustrations of diffusion (forward) processes on image, trajectories, and grasp poses (Urain et al. (2023)) a

Originality

Limitation & Further Study

Evaluation

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

์ดํ‰: ๋ณธ ๋…ผ๋ฌธ์€ ๋กœ๋ด‡ ์กฐ์ž‘ ๋ถ„์•ผ์—์„œ ๋น ๋ฅด๊ฒŒ ์„ฑ์žฅํ•˜๋Š” diffusion model ์—ฐ๊ตฌ๋ฅผ ์ฒ˜์Œ์œผ๋กœ ์ฒด๊ณ„์ ์œผ๋กœ ์ •๋ฆฌํ•œ ๊ฐ€์น˜ ์žˆ๋Š” survey๋กœ, ์—ฐ๊ตฌ์ž์™€ ์‹ค๋ฌด์ž ๋ชจ๋‘์—๊ฒŒ ํ•„์ˆ˜์ ์ธ ์ฐธ๊ณ ์ž๋ฃŒ๋ฅผ ์ œ๊ณตํ•œ๋‹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˜์ •