KOINEU Logo
No Image

Online Single-Channel Audio-Based Sound Speed Estimation for Robust Multi-Channel Audio Control

1๏ธโƒฃ ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๊ณต๊ฐ„ ์˜ค๋””์˜ค ์ œ์–ด(SZC, ANC ๋“ฑ) ๋Š” ์—ฌ๋Ÿฌ ์Šคํ”ผ์ปค์™€ ๋งˆ์ดํฌ๋ฅผ ํ™œ์šฉํ•ด ํŠน์ • ์˜์—ญ์— ์›ํ•˜๋Š” ์‚ฌ์šด๋“œ๋ฅผ ๋งŒ๋“ค๊ณ , ๋‹ค๋ฅธ ์˜์—ญ์—์„œ๋Š” ์†Œ์Œ์„ ์–ต์ œํ•œ๋‹ค. ์ด๋Ÿฌํ•œ ์‹œ์Šคํ…œ์€ ์‚ฌ์ „ ์ธก์ •๋œ ์ž„ํŽ„์Šค ์‘๋‹ต(IR) ์— ๊ธฐ๋ฐ˜ํ•ด ํ•„ํ„ฐ๋ฅผ ์„ค๊ณ„ํ•˜์ง€๋งŒ, ์ฒญ์ทจ์ž ์ด๋™ยทํŠธ๋žœ์Šค๋“€์„œ ๋“œ๋ฆฌํ”„ํŠธยท์˜จ๋„ยท์Šต๋„ ๋ณ€ํ™” ๋“ฑ ํ™˜๊ฒฝ ์š”์ธ์œผ๋กœ IR์ด ๋ณ€ํ•˜๋ฉด ์„ฑ๋Šฅ์ด ๊ธ‰๊ฒฉํžˆ ์ €ํ•˜๋œ๋‹ค. ํŠนํžˆ ์Œ์† ๋ณ€ํ™” ๋Š” ์ „ํŒŒ ์ง€์—ฐ๊ณผ ์œ„์ƒ์— ์ง์ ‘์ ์ธ ์˜ํ–ฅ์„ ์ฃผ์–ด, ๊ธฐ์กด ์ œ์–ด ํ•„ํ„ฐ๊ฐ€ ๊ธ‰๊ฒฉํžˆ ๋น„ํšจ์œจ์ ์œผ๋กœ ๋งŒ๋“ ๋‹ค. ๊ธฐ์กด ์—ฐ๊ตฌ๋“ค์€ ๋‹ค์ค‘ ๋งˆ์ดํฌ, ๋ณ„๋„ ์บ˜๋ฆฌ๋ธŒ๋ ˆ์ด์…˜, ์˜จยท์Šต๋„ ์„ผ์„œ ๋“ฑ์„ ์ „

Audio Processing Electrical Engineering and Systems Science
No Image

Open diffusion MRI and connectivity data for epilepsy and surgery: The IDEAS II release

1. ์—ฐ๊ตฌ์˜ ์˜์˜ ๋ฐ์ดํ„ฐ ๊ณต๋ฐฑ ๋ฉ”์šฐ๊ธฐ : ๊ธฐ์กด ๊ณต๊ฐœ ๊ฐ„์งˆ ์˜์ƒ ๋ฐ์ดํ„ฐ๋ฒ ์ด์Šค(์˜ˆ: Human Connectome, ADNI, ABIDE)์™€ ๋‹ฌ๋ฆฌ, ๊ฐ„์งˆ ํ™˜์ž์— ๋Œ€ํ•œ ๋Œ€๊ทœ๋ชจ DWI ๋ฐ์ดํ„ฐ๊ฐ€ ๊ฑฐ์˜ ์—†์—ˆ์Œ. IDEAS II๋Š” 216๋ช…์ด๋ผ๋Š” ๊ทœ๋ชจ์™€ ํ’๋ถ€ํ•œ ์ž„์ƒ ๋ฉ”ํƒ€๋ฐ์ดํ„ฐ๋ฅผ ๊ฒฐํ•ฉํ•ด, ๊ตฌ์กฐ์  ์—ฐ๊ฒฐ์„ฑ ์—ฐ๊ตฌ์™€ ์ˆ˜์ˆ  ๊ฒฐ๊ณผ ์˜ˆ์ธก ๋ชจ๋ธ๋ง์— ํ•„์ˆ˜์ ์ธ ์ž์›์„ ์ œ๊ณตํ•œ๋‹ค. ๋‹ค์ค‘ ํŒŒ์ดํ”„๋ผ์ธ ์ œ๊ณต : ์›์‹œยท์ „์ฒ˜๋ฆฌยทํŒŒ์ƒ ๋ฐ์ดํ„ฐ ๋ชจ๋‘๋ฅผ ๊ณต๊ฐœํ•จ์œผ๋กœ์จ, ์ดˆ๋ณด ์—ฐ๊ตฌ์ž๋Š” ๋ฐ”๋กœ ์ „์ฒ˜๋ฆฌ๋œ ๋ฐ์ดํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜๊ณ , ๊ณ ๊ธ‰ ์‚ฌ์šฉ์ž๋Š” ์ž์ฒด ํŒŒ์ดํ”„๋ผ์ธ์„ ์ ์šฉํ•ด ์žฌํ˜„์„ฑ์„ ๊ฒ€์ฆํ•  ์ˆ˜ ์žˆ๋‹ค. 2.

Data Quantitative Biology
No Image

Optical Inversion and Spectral Unmixing of Spectroscopic Photoacoustic Images with Physics-Informed Neural Networks

| ํ•ญ๋ชฉ | ๋‚ด์šฉ ๋ฐ ํ‰๊ฐ€ | | | | | ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ | sPA ์˜์ƒ์€ ๊นŠ์ด(์ˆ˜ cm)๊นŒ์ง€ ๊ณ ํ•ด์ƒ๋„ ๊ตฌ์กฐยท๊ธฐ๋Šฅ ์ •๋ณด๋ฅผ ์ œ๊ณตํ•˜์ง€๋งŒ, ํŒŒ์žฅโ€‘์˜์กด ํ”Œ๋ฃจ์–ธ์Šค์™€ ์‚ฐ๋ž€์œผ๋กœ ์ธํ•ด ๋น„์„ ํ˜•ยทillโ€‘posed ๋ฌธ์ œ์— ์ง๋ฉดํ•œ๋‹ค. ๊ธฐ์กด ์„ ํ˜• ๋ฐฉ๋ฒ•(NLS, NMF)์€ ๋น ๋ฅด์ง€๋งŒ ํ”Œ๋ฃจ์–ธ์Šค ๋ณด์ •์ด ๋ถ€์กฑํ•˜๊ณ , ์ตœ์‹  ๋น„์„ ํ˜• ML ๋ฐฉ๋ฒ•์€ ๋ผ๋ฒจ์ด ์žˆ๋Š” ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋ฐ์ดํ„ฐ ์— ์˜์กดํ•œ๋‹ค. ๋ผ๋ฒจ์ด ์—†๋Š” ์‹ค์ œ ๋ฐ์ดํ„ฐ๋ฅผ ์ง์ ‘ ํ•™์Šตํ•  ์ˆ˜ ์žˆ๋Š” ํ”„๋ ˆ์ž„์›Œํฌ๊ฐ€ ์ ˆ์‹คํžˆ ํ•„์š”ํ–ˆ๋‹ค. | | ํ•ต์‹ฌ ์•„์ด๋””์–ด | Physicsโ€‘Informed Neural Network (PINN)

Computer Science Network Machine Learning
No Image

Optimal Placement and Sizing of PV-Based DG Units in a Distribution Network Considering Loading Capacity

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์žฌ์ƒ์—๋„ˆ์ง€ยทDG ํ™•๋Œ€ ๋Š” ์ „๋ ฅ๊ณ„ํ†ต์˜ ์ €ํƒ„์†Œ ์ „ํ™˜์— ํ•„์ˆ˜์ ์ด์ง€๋งŒ, ๊ฐ„ํ—์„ฑ ๊ณผ ๋ฐฐ์ „๋ง ์ˆ˜์šฉ ํ•œ๊ณ„ ๋ฌธ์ œ๋กœ ์ธํ•ด ์ „์•• ์•ˆ์ •์„ฑยท์†์‹ค ๊ฐ์†Œ ๋“ฑ ์šด์˜์ƒ์˜ ์ œ์•ฝ์ด ํฌ๊ฒŒ ์ž‘์šฉํ•œ๋‹ค. ๊ธฐ์กด ์ตœ์  ๋ฐฐ์น˜ ์—ฐ๊ตฌ๋Š” ๋‹จ์ผ ๋ชฉํ‘œ(์†์‹ค ์ตœ์†Œํ™” ๋˜๋Š” ์ „์•• ๊ฐœ์„ ) ์— ์ง‘์ค‘ํ•˜๊ฑฐ๋‚˜, ์ œ์•ฝ ์กฐ๊ฑด(์„ ๋ฅ˜ ํ•œ๊ณ„ยท์ „์•• ํ•œ๊ณ„ ๋“ฑ) ์„ ์ถฉ๋ถ„ํžˆ ๋ฐ˜์˜ํ•˜์ง€ ๋ชปํ•˜๋Š” ๊ฒฝ์šฐ๊ฐ€ ๋งŽ๋‹ค. 2. ์ œ์•ˆ ๋ฐฉ๋ฒ•๋ก ์˜ ํ•ต์‹ฌ ์•„์ด๋””์–ด | ๋‹จ๊ณ„ | ๋ชฉํ‘œ | ์ฃผ์š” ๊ธฐ๋ฒ• | ๊ธฐ๋Œ€ ํšจ๊ณผ | | | | | | | 1๋‹จ๊ณ„ | ์ถ”๊ฐ€ ํ™œ์„ฑ ๋ถ€ํ•˜ ์šฉ๋Ÿ‰ (Loadability) ์‚ฐ์ถœ | ๋ฐ˜๋ณต์  ๋ถ€ํ•˜ ํ™•

Network Electrical Engineering and Systems Science
No Image

Optimized 3D Gaussian Splatting using Coarse-to-Fine Image Frequency Modulation

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ 3D Gaussian Splatting(3DGS) ์€ ์ตœ๊ทผ ์‹ ๊ทœ ์‹œ์  ํ•ฉ์„ฑ ์—์„œ ๋›ฐ์–ด๋‚œ ํ’ˆ์งˆ๊ณผ ์‹ค์‹œ๊ฐ„ ๋ Œ๋”๋ง์„ ์ œ๊ณตํ•˜์ง€๋งŒ, ์ˆ˜๋ฐฑ๋งŒ ๊ฐœ ์— ๋‹ฌํ•˜๋Š” ๊ฐ€์šฐ์‹œ์•ˆ ํ”„๋ฆฌ๋ฏธํ‹ฐ๋ธŒ๊ฐ€ ํ•„์š”ํ•ด GPU ๋ฉ”๋ชจ๋ฆฌยท๋””์Šคํฌ ์šฉ๋Ÿ‰ ์ด ํฌ๊ฒŒ ์†Œ๋ชจ๋œ๋‹ค. ์†Œ๋น„์ž์šฉ GPU(์˜ˆ: RTX 3060)์—์„œ๋Š” ๋ฉ”๋ชจ๋ฆฌ ํ•œ๊ณ„๊ฐ€ ๋นจ๋ฆฌ ๋„๋‹ฌํ•˜๊ณ , ๋Œ€๊ทœ๋ชจ ์”ฌ์„ ๋‹ค๋ฃจ๋Š” ๋ชจ๋ฐ”์ผยทAR/VR ๋””๋ฐ”์ด์Šค์—์„œ๋Š” ์ ์šฉ์ด ์–ด๋ ค์šด ์‹ค์ •์ด๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด Coarseโ€‘toโ€‘Fine ์ด๋ฏธ์ง€ ์ฃผํŒŒ์ˆ˜ ๋ชจ๋“ˆ๋ ˆ์ด์…˜ : ํ•™์Šต ์ดˆ๊ธฐ์— ์ €์ฃผํŒŒ(๊ฑฐ์นœ) ์ด๋ฏธ์ง€๋งŒ ์‚ฌ์šฉํ•ด ๊ฐ€์šฐ์‹œ์•ˆ์„ ํฌ๊ฒŒ ๋ฐฐ์น˜ํ•˜

No Image

Optimizing p-spin models through hypergraph neural networks and deep reinforcement learning

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ pโ€‘์Šคํ•€ ์œ ๋ฆฌ ๋Š” ๊ตฌ์กฐ์  ์œ ๋ฆฌ, ์ดˆ๋ƒ‰๊ฐ ์•ก์ฒด, ์œ„์ƒ ์–‘์ž ์˜ค๋ฅ˜ ์ •์ • ๋“ฑ ๋ฌผ๋ฆฌยทํ™”ํ•™ยท์–‘์ž ์ •๋ณด ๋ถ„์•ผ์—์„œ ํ•ต์‹ฌ ๋ชจ๋ธ์ด๋ฉฐ, p > 2์ผ ๋•Œ๋Š” ์ž์œ  ์—๋„ˆ์ง€ ์ง€ํ˜•์ด ๊ฑฐ์น ๊ณ  ํ”„๋ž™ํƒˆ ์ ์ธ ํŠน์„ฑ์„ ๋ณด์—ฌ ๊ธฐ์กด ํ‰๊ท ์žฅ ์ด๋ก ์ด ๋ถ•๊ดดํ•œ๋‹ค. ์ตœ์ ํ™”๋Š” NPโ€‘hard ์ด๋ฉฐ, ์ •ํ™• ์•Œ๊ณ ๋ฆฌ์ฆ˜์€ ์ˆ˜์‹ญ ์Šคํ•€ ์ˆ˜์ค€์— ๋จธ๋ฌผ๊ณ , ๋ฉ”ํƒ€ํœด๋ฆฌ์Šคํ‹ฑ(SA, PT ๋“ฑ)์€ ํƒ์ƒ‰ ํšจ์œจ์ด ๊ธ‰๊ฒฉํžˆ ์ €ํ•˜๋œ๋‹ค. ์ตœ๊ทผ ๋จธ์‹ ๋Ÿฌ๋‹ ๊ธฐ๋ฐ˜ ์ตœ์ ํ™” ๊ฐ€ ๋ถ€์ƒํ–ˆ์ง€๋งŒ, ๋Œ€๋ถ€๋ถ„์€ ๊ทธ๋ž˜ํ”„(2โ€‘์ฐจ) ๊ตฌ์กฐ ์— ๋งž์ถฐ ์„ค๊ณ„๋ผ ๊ณ ์ฐจ์› ์ƒํ˜ธ์ž‘์šฉ์„ ๋‹ค๋ฃจ๊ธฐ์—” ํ•œ๊ณ„๊ฐ€ ์žˆ๋‹ค. 2. ํ•ต์‹ฌ ๊ธฐ์—ฌ | ๋ฒˆํ˜ธ

Condensed Matter Network Learning Model
No Image

Orthogonal parametrisations of Extreme-Value distributions

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ | ๋ฌธ์ œ์  | ๊ธฐ์กด ์ ‘๊ทผ๋ฒ•์˜ ํ•œ๊ณ„ | | | | | ํ‘œ๋ณธ ํฌ๊ธฐ ์ œํ•œ โ€“ ๊ทน๊ฐ’ ๋ฐ์ดํ„ฐ๋Š” ๋ณดํ†ต ์ˆ˜์‹ญ~์ˆ˜๋ฐฑ ๊ฐœ์— ๋ถˆ๊ณผํ•จ | MLE๊ฐ€ ํŽธํ–ฅยท๋ถ„์‚ฐ์ด ํฌ๊ฒŒ ์ฆ๊ฐ€ | | ํŒŒ๋ผ๋ฏธํ„ฐ ์ƒ๊ด€ โ€“ ์œ„์น˜ยท๊ทœ๋ชจยทํ˜•์ƒ ํŒŒ๋ผ๋ฏธํ„ฐ๊ฐ€ ์„œ๋กœ ๊ฐ•ํ•˜๊ฒŒ ์–ฝํž˜ | ์ตœ์ ํ™”ยทMCMC์—์„œ ํ˜ผํ•ฉ ์†๋„ ์ €ํ•˜ | | ํ•ด์„์„ฑ ๋ถ€์กฑ โ€“ ํŒŒ๋ผ๋ฏธํ„ฐ ์ž์ฒด๊ฐ€ ์‹ค๋ฌด์—์„œ ์ง์ ‘์ ์ธ ์˜๋ฏธ๋ฅผ ๊ฐ–์ง€ ์•Š์Œ | ์ •์ฑ…ยท์œ„ํ—˜ ๊ด€๋ฆฌ์— ํ™œ์šฉ ์–ด๋ ค์›€ | Coxโ€‘Reid(1987)์˜ ์ •๊ทœ ์ง๊ต ํŒŒ๋ผ๋ฏธํ„ฐํ™” ๋Š” ํ”ผ์…” ์ •๋ณด ํ–‰๋ ฌ์˜ ๊ต์ฐจํ•ญ์„ 0์œผ๋กœ ๋งŒ๋“ค์–ด โ€œ๊ด€์‹ฌ ํŒŒ๋ผ๋ฏธํ„ฐ(ฯˆ)โ€์™€ โ€œ๋ณด์กฐ ํŒŒ๋ผ๋ฏธํ„ฐ(ฮป

Mathematics
No Image

Out-of-equilibrium selection pressure enhances inference from protein sequence data

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ์งˆ๋ฌธ ๊ณต๋™ ์ง„ํ™”์™€ ์ƒ๊ด€๊ด€๊ณ„ : ๋™์ผํ•œ ๊ตฌ์กฐยท๊ธฐ๋Šฅ์„ ์œ ์ง€ํ•ด์•ผ ํ•˜๋Š” ๋‹จ๋ฐฑ์งˆ์€ ์ ‘์ด‰ ๋ถ€์œ„์˜ ์•„๋ฏธ๋…ธ์‚ฐ์ด ์ƒํ˜ธ ๋ณด์™„์ ์ธ ๋ณ€ํ™”๋ฅผ ๊ฒช์œผ๋ฉฐ, ์ด๋Š” MSA์—์„œ ์ƒ๊ด€๊ด€๊ณ„(์ฝ”์—๋ณผ๋ฃจ์…˜)๋กœ ๋‚˜ํƒ€๋‚œ๋‹ค. ์ด๋Ÿฌํ•œ ์ฝ”์—๋ณผ๋ฃจ์…˜์€ Potts ๋ชจ๋ธ, Mutual Information, ์ตœ๊ทผ์˜ Protein Language Model ๋“ฑ ๋‹ค์–‘ํ•œ ๋ฐฉ๋ฒ•์œผ๋กœ ๊ตฌ์กฐยท๊ธฐ๋Šฅ์„ ์ถ”๋ก ํ•˜๋Š” ๊ทผ๊ฑฐ๊ฐ€ ๋œ๋‹ค. ์„ ํƒ์•• ๋ณ€๋™์˜ ๋ฏธ์ง€ : ์ž์—ฐ ํ™˜๊ฒฝ์€ ์‹œ์‹œ๊ฐ๊ฐ ๋ณ€ํ•˜๊ณ , ์ด์— ๋”ฐ๋ผ ๋‹จ๋ฐฑ์งˆ์— ๊ฐ€ํ•ด์ง€๋Š” ์„ ํƒ์••๋„ ๋ณ€ํ•œ๋‹ค. ๊ธฐ์กด ์—ฐ๊ตฌ๋Š” ์ฃผ๋กœ ํ‰ํ˜•(steadyโ€‘state) ๊ฐ€์ •์„ ์ „์ œ๋กœ

Data Quantitative Biology
No Image

Oxi-Shapes: Tropical geometric analysis of bounded redox proteomic state spaces

** ๋ ˆ๋…์Šค ํ”„๋กœํ…Œ์˜ค๋ฏน์Šค๋Š” 0โ€’1 ๊ตฌ๊ฐ„์œผ๋กœ ์ œํ•œ๋œ ์‚ฐํ™” ์ ์œ ์œจ์„ ์ œ๊ณตํ•˜์ง€๋งŒ, ๊ธฐ์กด์˜ ์„ ํ˜• ๋Œ€์ˆ˜์  ๋ถ„์„๋ฒ•์€ ์ด๋Ÿฌํ•œ ์ œํ•œ๋œ ์ƒํƒœ๊ณต๊ฐ„๊ณผ ๊ทผ๋ณธ์ ์œผ๋กœ ๋งž์ง€ ์•Š๋Š”๋‹ค. ๋ณธ ์—ฐ๊ตฌ๋Š” โ€˜Oxiโ€‘Shapesโ€™๋ผ๋Š” ์—ด๋Œ€๊ธฐํ•˜ํ•™(tropical geometry) ๊ธฐ๋ฐ˜ ํ”„๋ ˆ์ž„์›Œํฌ๋ฅผ ์ œ์•ˆํ•œ๋‹ค. Oxiโ€‘Shapes๋Š” ๊ฐ ์‹œ์Šคํ…Œ์ธ ์ž”๊ธฐ๋ฅผ ๋ถˆ๋ณ€์˜ ๊ฒฉ์ž์ ์œผ๋กœ ๋‘๊ณ , ์ธก์ •๋œ ์‚ฐํ™” ์ ์œ ์œจ์„ ๊ฒฉ์ž ์œ„์˜ ์Šค์นผ๋ผ ํ•„๋“œ๋กœ ํ‘œํ˜„ํ•œ๋‹ค. ์ด๋ฅผ ํ†ตํ•ด (1) ์ „์—ญ ์ˆ˜์ค€์—์„œ๋Š” ๋‚ด๋ถ€ ๋ ˆ๋…์Šค ์—”ํŠธ๋กœํ”ผ, ๊ฒฉ์ž ๊ณก๋ฅ , ํŒŒ์ƒ ์—๋„ˆ์ง€ ํ•จ์ˆ˜ ๋“ฑ์„ ์ •์˜ํ•ด ๋ ˆ๋…์Šค ํ”„๋กœํ…Œ์˜ค๋ฏน์Šค์˜ ๊ธฐํ•˜ํ•™์  ๊ตฌ์กฐ๋ฅผ ์ •๋Ÿ‰ํ™”

Quantitative Biology Analysis
No Image

Partial Identification under Missing Data Using Weak Shadow Variables from Pretrained Models

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ MNAR ๊ฒฐ์ธก ์€ ์„ค๋ฌธยท๊ฑด๊ฐ• ์กฐ์‚ฌยท๋””์ง€ํ„ธ ํ”Œ๋žซํผ ๋“ฑ์—์„œ ํ”ํžˆ ๋ฐœ์ƒํ•œ๋‹ค. ๊ฒฐ์ธก ํ™•๋ฅ ์ด ์‹ค์ œ ๊ฐ’์— ์˜์กดํ•˜๋ฏ€๋กœ ๊ด€์ธก ํ‰๊ท ์€ ํŽธํ–ฅ๋œ๋‹ค. ๊ธฐ์กด ํ•ด๊ฒฐ์ฑ…์€ (i) ํŒŒ๋ผ๋ฉ”ํŠธ๋ฆญ ์„ ํƒ ๋ชจ๋ธ(Heckman) , (ii) ํŒจํ„ดโ€‘๋ฏน์Šค์ฒ˜ ๋ชจ๋ธ , (iii) ๋ณด์กฐ๋ณ€์ˆ˜/๊ทธ๋ฆผ์ž ๋ณ€์ˆ˜ ๋“ฑ์ด๋‹ค. ํ•˜์ง€๋งŒ (i)ยท(ii)๋Š” ๊ฐ•ํ•œ ๊ตฌ์กฐ ๊ฐ€์ •์ด ํ•„์š”ํ•˜๊ณ , (iii)๋Š” ์™„์ „์„ฑ(completeness) ํ˜น์€ ๊ฐ•ํ•œ ๊ด€๋ จ์„ฑ(strong relevance) ์„ ์š”๊ตฌํ•œ๋‹ค. ์‹ค์ œ ๋ฐ์ดํ„ฐ์—์„œ๋Š” ์ด๋Ÿฌํ•œ ๋ณด์กฐ๋ณ€์ˆ˜๋ฅผ ์ฐพ๊ธฐ ์–ด๋ ต๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด: ์•ฝํ•œ ๊ทธ๋ฆผ

Data Statistics Machine Learning Model
No Image

Particle-in-Cell Methods for Simulations of Sheared, Expanding, or Escaping Astrophysical Plasma

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์™„์ „ ๋™์—ญํ•™ PIC ์€ ๋น„์ถฉ๋Œ์„ฑ ํ”Œ๋ผ์ฆˆ๋งˆ์˜ ๋ฏธ์„ธ ๋ฌผ๋ฆฌ(์ž…์ž ๊ฐ€์†, ํŒŒ๋™โ€‘์ž…์ž ์ƒํ˜ธ์ž‘์šฉ ๋“ฑ)๋ฅผ ์ง์ ‘ ํ•ด๊ฒฐํ•  ์ˆ˜ ์žˆ์–ด ์ฒœ์ฒด ๋ฌผ๋ฆฌํ•™์—์„œ ํ•ต์‹ฌ ๋„๊ตฌ๊ฐ€ ๋˜์—ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์ „ํ†ต์ ์ธ PIC ์€ ๊ท ์ผํ•˜๊ณ  ์ •์ ์ธ ๋ฐฐ๊ฒฝ์„ ๊ฐ€์ •ํ•œ๋‹ค. ์‹ค์ œ ์ฒœ์ฒด ํ™˜๊ฒฝ(์˜ˆ: ์›๋ฐ˜์˜ ์ฐจ๋™ ํšŒ์ „, ํƒœ์–‘ํ’ ํŒฝ์ฐฝ, ๊ณ ์—๋„ˆ์ง€ ๊ฐ€์†๊ธฐ ์ฃผ๋ณ€)์—์„œ๋Š” ์ „๋‹จ, ํŒฝ์ฐฝ/์ˆ˜์ถ•, ์ž…์ž ํƒˆ์ถœ ๊ณผ ๊ฐ™์€ ๊ฑฐ์‹œ์  ๋ณ€ํ˜•์ด ๋™์‹œ์— ์ผ์–ด๋‚˜๋ฉฐ, ์ด๋ฅผ ๋ฌด์‹œํ•˜๋ฉด ๋ฌผ๋ฆฌ์  ๊ฒฐ๊ณผ๊ฐ€ ํฌ๊ฒŒ ์™œ๊ณก๋œ๋‹ค. ๋”ฐ๋ผ์„œ ์ „๋‹จโ€‘๋ฐ•์Šค, ํŒฝ์ฐฝโ€‘๋ฐ•์Šค, ๋ˆ„์ˆ˜โ€‘๋ฐ•์Šค ์™€ ๊ฐ™์€ ๋กœ์ปฌ ๋ฐ•์Šค ๊ธฐ๋ฒ•์„ PIC์— ํ†ตํ•ฉํ•˜๋Š” ๊ฒƒ์ด ํ•„์ˆ˜

Physics
No Image

Passive Imaging with Ambient Noise Under Wave Speed Mismatch: Mathematical Analysis and Wave Speed Estimation

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ์ˆ˜๋™ ์ƒ๊ด€ ๊ธฐ๋ฐ˜ ์ด๋ฏธ์ง•(passive correlationโ€‘based imaging)์€ ์ธ์œ„์ ์ธ ์†Œ์Šค ์—†์ด๋„ ์žก์Œ์œผ๋กœ๋ถ€ํ„ฐ Green ํ•จ์ˆ˜์™€ ์ „ํŒŒ ์‹œ๊ฐ„์„ ๋ณต์›ํ•  ์ˆ˜ ์žˆ์–ด, ์ง€๊ตฌ๋ฌผ๋ฆฌยท์ฒœ๋ฌธยท๊ณตํ•™ ๋“ฑ ๋‹ค์–‘ํ•œ ๋ถ„์•ผ์—์„œ ๊ธ‰๋ถ€์ƒํ•˜๊ณ  ์žˆ๋‹ค. ๊ธฐ์กด ์—ฐ๊ตฌ๋Š” ์ฃผ๋กœ ํŒŒ๋™ ์†๋„๊ฐ€ ์ •ํ™•ํžˆ ์•Œ๋ ค์ง„ ๊ฒฝ์šฐ ์— ์ดˆ์ ์„ ๋งž์ถ”์—ˆ์œผ๋ฉฐ, ์†๋„ ๋ถˆ์ผ์น˜๊ฐ€ ์ด๋ฏธ์ง€ ํ’ˆ์งˆ์— ๋ฏธ์น˜๋Š” ์ •๋Ÿ‰์  ์˜ํ–ฅ์„ ์ฒด๊ณ„์ ์œผ๋กœ ๋‹ค๋ฃจ์ง€๋Š” ๋ชปํ–ˆ๋‹ค. ๋ณธ ๋…ผ๋ฌธ์€ ์ด๋Ÿฌํ•œ ๊ณต๋ฐฑ์„ ๋ฉ”์šฐ๊ธฐ ์œ„ํ•ด โ€œdaylightโ€ ๊ตฌ์„ฑ (์„ผ์„œ ๋ฐฐ์—ด์ด ์žก์Œ์›๊ณผ ๋งค์งˆ ์‚ฌ์ด์— ์œ„์น˜)์—์„œ ์†๋„ ๋ถˆ์ผ์น˜ ๊ฐ€ ์ด๋ฏธ์ง€์—

Electrical Engineering and Systems Science Analysis
No Image

Perceptive Humanoid Parkour: Chaining Dynamic Human Skills via Motion Matching

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ํœด๋จธ๋…ธ์ด๋“œ ํŒŒ์ฟ ๋ฅด ๋Š” ๊ณ ๋‹ค์ด๋‚ด๋ฏนยท์ ‘์ด‰โ€‘๋‹ค์–‘ํ•œ ๋™์ž‘๊ณผ ์‹ค์‹œ๊ฐ„ ํ™˜๊ฒฝ ์ธ์‹์ด ๋™์‹œ์— ์š”๊ตฌ๋˜๋Š” ๊ฐ€์žฅ ๊นŒ๋‹ค๋กœ์šด ํ…Œ์ŠคํŠธ๋ฒ ๋“œ์ด๋‹ค. ๊ธฐ์กด ์—ฐ๊ตฌ๋Š” ์ฃผ๋กœ ๋ณดํ–‰ยท๊ณ„๋‹จยท๋‚™์„ ๋“ฑ ์ €๋‹ค์ด๋‚ด๋ฏน ์ง€ํ˜•์— ์ดˆ์ ์„ ๋งž์ถ”์—ˆ์œผ๋ฉฐ, ์ธ๊ฐ„ ๋™์ž‘ ๋ฐ์ดํ„ฐ๋ฅผ ํ™œ์šฉํ•œ ๊ณ ์† ์Šคํ‚ฌ์€ ๋ฐ์ดํ„ฐ ๋ถ€์กฑ๊ณผ ์ „์ด ๋ฌธ์ œ ๋•Œ๋ฌธ์— ์ œํ•œ์ ์ด์—ˆ๋‹ค. 2. ํ•ต์‹ฌ ๊ธฐ์—ฌ (Contributions) | ๋ฒˆํ˜ธ | ๋‚ด์šฉ | ์˜์˜ | | | | | | 1 | ๋ชจ์…˜ ๋งค์นญ ๊ธฐ๋ฐ˜ ์žฅ๊ธฐ ์Šคํ‚ฌ ์ฒด์ธ ํŒŒ์ดํ”„๋ผ์ธ โ€“ ์›์ž ์Šคํ‚ฌ์„ locomotion๊ณผ ์—ฐ๊ฒฐํ•ด ๋‹ค์–‘ํ•œ ์ ‘๊ทผยท์ „ํ™˜ ์ƒํ™ฉ์„ ์ž๋™ ์ƒ์„ฑ | ํฌ์†Œํ•œ

Robotics Computer Science
No Image

Phase Transitions in Collective Damage of Civil Structures under Natural Hazards

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ๋„์‹œ ๊ทœ๋ชจ ์œ„ํ—˜์˜ ๋ณตํ•ฉ์„ฑ : 2023๋…„ ์นดํ๋ผ๋งŒ๋ผ์‹œ ์ง€์ง„ยท2024๋…„ ํ—ˆ๋ฆฌ์ผ€์ธ ๋“ฑ ์ตœ๊ทผ ์žฌ๋‚œ ์‚ฌ๋ก€๋Š” ๊ฐœ๋ณ„ ๊ฑด๋ฌผ ์†์ƒ์„ ๋„˜์–ด ๋„์‹œ ์ „์ฒด์˜ ๊ธฐ๋Šฅ ๋งˆ๋น„์™€ ๊ฒฝ์ œ์  ์†์‹ค์„ ๊ฐ•์กฐํ•œ๋‹ค. ์ง‘๋‹จ์  ์†์ƒ์˜ ๋ฏธ์ง€ ์˜์—ญ : ๊ธฐ์กด ์žฌ๋‚œ ์œ„ํ—˜ ๋ชจ๋ธ์€ ์ฃผ๋กœ ๊ฑด๋ฌผ ๊ฐœ๋ณ„์˜ ๋‚ด์ง„ยท๋‚ดํ’ ์„ฑ๋Šฅ์— ์ดˆ์ ์„ ๋งž์ถ”์—ˆ์œผ๋ฉฐ, ๊ฑด๋ฌผ ๊ฐ„ ์ƒํ˜ธ์ž‘์šฉ์ด๋‚˜ ํฌํŠธํด๋ฆฌ์˜ค ์ „์ฒด์˜ ํ†ต๊ณ„์  ๊ฑฐ๋™์€ ์ถฉ๋ถ„ํžˆ ํƒ๊ตฌ๋˜์ง€ ์•Š์•˜๋‹ค. 2. ํ•ต์‹ฌ ๋ฐฉ๋ฒ•๋ก  | ๋‹จ๊ณ„ | ๋‚ด์šฉ | ์ฃผ์š” ๊ธฐ์ˆ ยท๋„๊ตฌ | | | | | | ๋ฐ์ดํ„ฐ ์ˆ˜์ง‘ | Milpitas(์บ˜๋ฆฌํฌ๋‹ˆ์•„) ๋‹ค์ธต ๊ฑด๋ฌผ ์žฌ๊ณ  ํ™•๋ณด | ๊ฑด๋ฌผ ์—ฐ

Statistics Applications
No Image

Phase-Based Bit Commitment Protocol

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ํ”„๋ผ์ด๋ฒ„์‹œ ์œ„๊ธฐ : AIยทML ์„œ๋น„์Šค์— ๋Œ€๊ทœ๋ชจ ๊ฐœ์ธ ๋ฐ์ดํ„ฐ๊ฐ€ ์œ ์ž…๋˜๋ฉด์„œ, ๋ฐ์ดํ„ฐ ์ œ๊ณต์ž๊ฐ€ ์ง์ ‘ ๋ฐ์ดํ„ฐ๋ฅผ ๋…ธ์ถœํ•˜์ง€ ์•Š๊ณ ๋„ ์—ฐ์‚ฐ์„ ์ˆ˜ํ–‰ํ•  ์ˆ˜ ์žˆ๋Š” Secure Function Evaluation (SFE) ๊ฐ€ ํ•„์ˆ˜์ ์ด๋‹ค. ๋น„ํŠธ ์ปค๋ฐ‹์˜ ํ•ต์‹ฌ์„ฑ : BC๋Š” SFE, ์˜์ง€์‹ ์ฆ๋ช…, ์ „์ž ํˆฌํ‘œ ๋“ฑ ๋‹ค์–‘ํ•œ ๊ณ ์œ„ ํ”„๋กœํ† ์ฝœ์˜ ๊ธฐ๋ณธ ๋นŒ๋”ฉ ๋ธ”๋ก์ด๋ฉฐ, ์–‘์ž ์—ญํ•™ ํ•˜์—์„œ๋Š” noโ€‘go ์ •๋ฆฌ (Mayersโ€‘Loโ€‘Chau) ๋•Œ๋ฌธ์— ๋ถˆ๊ฐ€๋Šฅํ•˜๋‹ค๊ณ  ์•Œ๋ ค์ ธ ์žˆ๋‹ค. ๊ฐ€์ • ๊ธฐ๋ฐ˜ ์ ‘๊ทผ : โ€œ๋„คํŠธ์›Œํฌ ์ œ๊ณต์ž๊ฐ€ ์ „์†ก์„ ๋กœ๋ฅผ ๋ฌผ๋ฆฌ์ ์œผ๋กœ ๋ณดํ˜ธํ•œ๋‹คโ€๋Š” Trus

Computer Science Cryptography and Security
No Image

Physical Activity Trajectories Preceding Incident Major Depressive Disorder Diagnosis Using Consumer Wearable Devices in the All of Us Research Program: Case-Control Study

1. ์—ฐ๊ตฌ ์„ค๊ณ„ ๋ฐ ๋ฐ์ดํ„ฐ ๊ฐ•์  | ์š”์†Œ | ์„ค๋ช… | ์˜์˜ | | | | | | ๋ฐ์ดํ„ฐ ์ถœ์ฒ˜ | All of Us ์—ฐ๊ตฌ ํ”„๋กœ๊ทธ๋žจ โ€“ EHR + Fitbit ์—ฐ๋™ | ๋Œ€๊ทœ๋ชจ(>30 k) ์‹ค์ƒํ™œ ๋ฐ์ดํ„ฐ, ๊ตญ๊ฐ€๋Œ€ํ‘œ ํ‘œ๋ณธ | | ์‚ฌ๊ฑดโ€‘๋Œ€์กฐ ์„ค๊ณ„ | ์ง„๋‹จ ์‹œ์  ๊ธฐ์ค€ 1:4 ๋งค์นญ(์—ฐ๋ นยฑ1 yr, ์„ฑ๋ณ„, BMI) | ํ˜ผ๋ž€ ๋ณ€์ˆ˜ ์ตœ์†Œํ™”, ํ†ต๊ณ„ ํšจ์œจ์„ฑ ํ™•๋ณด | | PA ์ธก์ • | ์ผ์ผ ๊ฑธ์Œ ์ˆ˜ + MVPA(โ€œfairly activeโ€+2ร—โ€œvery activeโ€) | ๊ฐ๊ด€์ , ์—ฐ์†์ , ์ผ์ƒ ์ƒํ™œ ๋ฐ˜์˜ | | ๋ฐ์ดํ„ฐ ํ’ˆ์งˆ | ํ•˜๋ฃจ 10 h ์ฐฉ์šฉ, 10

Statistics Applications
No Image

Physics-Informed Anomaly Detection of Terrain Material Change in Radar Imagery

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์ง€ํ˜• ๋ฌผ์งˆ ๋ณ€ํ™” ํƒ์ง€ ๋Š” ์ธํ”„๋ผ ๋ชจ๋‹ˆํ„ฐ๋ง, ํ™˜๊ฒฝ ๊ฐ์‹œ, ์•ˆ๋ณด ๋“ฑ ์‹ค์šฉ์  ์‘์šฉ ๋ถ„์•ผ์—์„œ ํ•ต์‹ฌ ๊ณผ์ œ์ด๋‹ค. ๊ธฐ์กด ๋ ˆ์ด๋” ๊ธฐ๋ฐ˜ ๋ณ€ํ™” ํƒ์ง€๋Š” ๊ฐ•๋„ ์ฐจ์ด ํ˜น์€ interferometric coherence ์— ์˜์กดํ•˜์ง€๋งŒ, ๋ฌผ๋ฆฌ์  ๋งค๊ฐœ๋ณ€์ˆ˜(์œ ์ „์œจยท๊ฑฐ์น ๊ธฐยท์ˆ˜๋ถ„)์™€์˜ ์ง์ ‘์ ์ธ ์—ฐ๊ด€์„ฑ์„ ์ถฉ๋ถ„ํžˆ ๋ฐ˜์˜ํ•˜์ง€ ๋ชปํ•œ๋‹ค. ๋˜ํ•œ, ๋ ˆ์ด๋” ํด๋Ÿฌํ„ฐ๊ฐ€ heavyโ€‘tailed (Gamma, Kโ€‘๋ถ„ํฌ) ํŠน์„ฑ์„ ๋ณด์ด๊ธฐ ๋•Œ๋ฌธ์— ์ „ํ†ต์ ์ธ Gaussian ๊ธฐ๋ฐ˜ RX ํƒ์ง€๋Š” CFAR(Constant False Alarm Rate) ์œ ์ง€๊ฐ€ ์–ด๋ ต๋‹ค. 2.

Electrical Engineering and Systems Science Detection
No Image

Piecewise integrability of the discrete Hasimoto map for analytic prediction and design of helical peptides

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ์ ๋ถ„๊ฐ€๋Šฅ๊ณ„์™€ ์ƒ๋ฌผํ•™ : NLS์™€ ๊ทธ ์ด์‚ฐํ˜•(DNLS)์€ ๊ด‘์„ฌ์œ , BEC ๋“ฑ ๋ฌผ๋ฆฌ๊ณ„์—์„œ ์ •ํ™•ํ•œ ์†”๋ฆฌํ†ค ํ•ด๋ฅผ ์ œ๊ณตํ•œ๋‹ค. ์ด๋ฅผ ๋‹จ๋ฐฑ์งˆ ๊ณจ๊ฒฉ์— ์ ์šฉํ•˜๋ ค๋Š” ์‹œ๋„๋Š” ํ”„๋ ˆ๋„ฌโ€‘์†”๋ฆฌํ†ค ์ „ํ†ต (Danielsson, Chernodub ๋“ฑ)์—์„œ ์‹œ์ž‘๋˜์—ˆ์œผ๋ฉฐ, ฮฑโ€‘ํ—ฌ๋ฆญ์Šค๋ฅผ ์†”๋ฆฌํ†ค ํฅ๋ถ„์œผ๋กœ ํ•ด์„ํ•œ๋‹ค. ์ „์—ญ ์ ๋ถ„๊ฐ€๋Šฅ์„ฑ์˜ ํ•œ๊ณ„ : ์žฅ๊ฑฐ๋ฆฌ ์ƒํ˜ธ์ž‘์šฉ, ํ‚ค๋ž„ ์ œ์•ฝ, ์„œ์—ด ํŠน์ด์  ํ™”ํ•™์„ฑ ๋“ฑ์œผ๋กœ ์ธํ•ด ์ „์ฒด ๋‹จ๋ฐฑ์งˆ์— ๋Œ€ํ•œ ์ „์—ญ DNLS ์ ์šฉ์€ ๋™์—ญํ•™์  ์˜ˆ์ธก ๋ณด๋‹ค๋Š” ๊ธฐํ•˜ํ•™์  ์„œ์ˆ  ์— ๋จธ๋ฌธ๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด โ€“ ๊ตฌ๊ฐ„์  ์ ๋ถ„๊ฐ€๋Šฅ์„ฑ (Piecewi

Quantitative Biology
No Image

Piezoelectric MEMS Phase Modulator for Silicon Nitride Platform in the Visible Spectrum

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๊ฐ€์‹œ๊ด‘์„  PIC : ๋ฐ”์ด์˜ค์ด๋ฏธ์ง•, ๋ผ์ด๋‹ค, ์–‘์ž ๊ธฐ์ˆ  ๋“ฑ์—์„œ ์ €์†์‹คยท๋„“์€ ํˆฌ๊ณผ์„ฑ์„ ๊ฐ–๋Š” Siโ‚ƒNโ‚„ ํŒŒํ˜• ๊ฐ€์ด๋“œ๋Š” ๊ฐ€์žฅ ์œ ๋งํ•œ ํ”Œ๋žซํผ์ด์ง€๋งŒ, ๋ณธ๋ž˜๋Š” ์ „๊ธฐยท์—ดยท๊ธฐ๊ณ„์  ๋ณ€์กฐ ํšจ์œจ์ด ๋‚ฎ์•„ ํ™œ์„ฑํ™”๊ฐ€ ์–ด๋ ค์› ๋‹ค. ๊ธฐ์กด ๋ณ€์กฐ ๊ธฐ์ˆ  ํ•œ๊ณ„ Thermoโ€‘optic : ์ „๋ ฅ ์†Œ๋ชจ(mW ์ˆ˜์ค€)์™€ ์—ด ํฌ๋กœ์Šคํ† ํฌ๊ฐ€ ํฌ๋ฉฐ, ์†๋„ ์ œํ•œ์ด ์žˆ๋‹ค. Pockels EO : ๋น ๋ฅด์ง€๋งŒ PZTยทLiNbOโ‚ƒยทZnO ๋“ฑ ์™ธ๋ถ€ ๋ฐ•๋ง‰์„ ๋„์ž…ํ•˜๋ฉด ๊ด‘์†์‹ค์ด ํฌ๊ฒŒ ์ฆ๊ฐ€ํ•œ๋‹ค. Stressโ€‘optic : PZT๋ฅผ ์ด์šฉํ•ด ์ŠคํŠธ๋ ˆ์Šค๋ฅผ ๊ฐ€ํ•˜๋ฉด Vฯ€ยทL โ‰ˆ 16 Vยทc

Physics
No Image

Pinching Antennas-Aided Integrated Sensing and Multicast Communication Systems

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ISAC ์€ ์ŠคํŽ™ํŠธ๋Ÿผ ํšจ์œจ๊ณผ ์‹œ์Šคํ…œ ํšจ์œจ์„ ๋™์‹œ์— ๋†’์ด๊ธฐ ์œ„ํ•ด ํ†ต์‹ ยท๊ฐ์ง€๋ฅผ ํ•˜๋‚˜์˜ ๋ฌด์„  ์ž์›์— ํ†ตํ•ฉํ•œ๋‹ค๋Š” ์ ์—์„œ ํ•ต์‹ฌ ๊ธฐ์ˆ ๋กœ ๋ถ€์ƒํ•˜๊ณ  ์žˆ๋‹ค. ๊ธฐ์กด MIMO ๊ธฐ๋ฐ˜ ISAC๋Š” ๊ณ ์ •๋œ ์•ˆํ…Œ๋‚˜ ๋ฐฐ์—ด์— ์˜์กดํ•ด ๋™์  ์ „ํŒŒ ํ™˜๊ฒฝ์— ๋Œ€ํ•œ ์ ์‘์„ฑ์ด ์ œํ•œ๋œ๋‹ค. ์œ ์—ฐ ์•ˆํ…Œ๋‚˜ (movable, fluid)์™€ RIS ๋Š” ๊ณต๊ฐ„ ์ ์‘์„ฑ์„ ์ œ๊ณตํ•˜์ง€๋งŒ ์ด๋™ ๋ฒ”์œ„ยท์ด์ค‘ ๊ฒฝ๋กœ ์†์‹ค ๋“ฑ ๋ฌผ๋ฆฌ์  ํ•œ๊ณ„๊ฐ€ ์žˆ๋‹ค. PASS ๋Š” ํŒŒ์žฅ ๊ฐ€์ด๋“œ ๋‚ด๋ถ€์— ํ•€์นญ ์•ˆํ…Œ๋‚˜๋ฅผ ์ž์œ ๋กญ๊ฒŒ ๋ถ€์ฐฉยท๋ถ„๋ฆฌํ•จ์œผ๋กœ์จ, ๊ธด ํŒŒ์žฅ ๊ฐ€์ด๋“œ๋ฅผ ํ†ตํ•œ โ€œnearโ€‘wiredโ€ ๊ฐ•ํ•œ LOS ๋ง

System Electrical Engineering and Systems Science
No Image

Planar Structures of Medium-Sized Gold Clusters Become Ground States upon Ionization

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ์†Œํ˜• ๊ธˆ ํด๋Ÿฌ์Šคํ„ฐ (n โ‰ค 13)๋Š” ์ „์ž ๊ตฌ์กฐ์™€ relativistic ํšจ๊ณผ ๋•Œ๋ฌธ์— 2D ๊ตฌ์กฐ๊ฐ€ ์„ ํ˜ธ๋˜๋Š” ๊ฒƒ์ด ์ž˜ ์•Œ๋ ค์ ธ ์žˆ๋‹ค. ์ค‘ยท๋Œ€ํ˜• ํด๋Ÿฌ์Šคํ„ฐ (n โ‰ˆ 20โ€“100)๋Š” 3D icosahedral, decahedral, octahedral ๋“ฑ ๋‹ค์–‘ํ•œ ๊ตฌ์กฐ๊ฐ€ ๊ฑฐ์˜ ๋™๋“ฑํ•œ ์—๋„ˆ์ง€๋ฅผ ๊ฐ€์ ธ, ์›์ž ํ•˜๋‚˜ ์ถ”๊ฐ€๋งŒ์œผ๋กœ๋„ ์ตœ์ €๊ตฌ์กฐ๊ฐ€ ๋ฐ”๋€Œ๋Š” ๋ณต์žกํ•œ PES๋ฅผ ๋ณด์ธ๋‹ค. ์ตœ๊ทผ 2D ๊ธˆ ์›์ž๋ง‰ (์ž์œ  ์ž…์ž์ธต, ๊ทธ๋ž˜ํ•€ ์•„๋ž˜ ์‚ฝ์ž… ๋“ฑ)์˜ ์‹คํ—˜์  ์‹คํ˜„ ๊ฐ€๋Šฅ์„ฑ์ด ์ œ์‹œ๋˜๋ฉด์„œ, โ€œ์–‘์ด์˜จํ™”๊ฐ€ 2D ๊ตฌ์กฐ๋ฅผ ์•ˆ์ •ํ™”ํ•  ์ˆ˜ ์žˆ๋‹คโ€๋Š” ๊ฐ€์„ค์ด ๋“ฑ์žฅํ–ˆ๋‹ค

Condensed Matter
No Image

Ponomarenko dynamo sustained by a free swirling jet

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ์˜์˜ ํฌ๋…ธ๋งˆ๋ Œ์ฝ” ๋‹ค์ด๋„ˆ๋ชจ ๋Š” ๊ณ ์ „์ ์ธ ๋‚˜์„ ํ˜• ํ๋ฆ„์ด ์ž๊ธฐ์žฅ์„ ์ƒ์„ฑํ•  ์ˆ˜ ์žˆ์Œ์„ ์ตœ์ดˆ๋กœ ์ œ์‹œํ•œ ๋ชจ๋ธ์ด๋ฉฐ, ๋ผ๊ฐ€ ์‹คํ—˜์—์„œ ์‹ค์ฆ๋˜์—ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๋ผ๊ฐ€์™€ ์นด๋ฅผ์Šค๋ฃจ์— ์‹คํ—˜์€ ๋‚ด๋ถ€ ๋ฒฝ๊ณผ ๊ฐ•์ œ๋œ ํ๋ฆ„ ์— ํฌ๊ฒŒ ์˜์กดํ•ด ๋น„์„ ํ˜• ์ƒํ˜ธ์ž‘์šฉ์„ ์ œํ•œํ–ˆ๋‹ค. ๋ณธ ์—ฐ๊ตฌ๋Š” ๋‚ด๋ถ€ ๊ตฌ์กฐ๊ฐ€ ์ „ํ˜€ ์—†๋Š” ์›ํ†ต ์„ ์ด์šฉํ•ด ํ๋ฆ„์ด ์ž์œ ๋กญ๊ฒŒ ๋ฐ˜์‘ํ•˜๋„๋ก ํ•จ์œผ๋กœ์จ, ๋น„์„ ํ˜• ์—ญํ•™ (ํฌํ™”, ์ง„๋™, ํ˜ผ๋ˆ ๋“ฑ)์„ ๋ณด๋‹ค ์ž์—ฐ์Šค๋Ÿฝ๊ฒŒ ํƒ๊ตฌํ•  ์ˆ˜ ์žˆ๋Š” ํ”Œ๋žซํผ์„ ์ œ์‹œํ•œ๋‹ค. 2. ์ˆ˜์น˜ ๋ชจ๋ธ๋ง ์ ‘๊ทผ๋ฒ• | ๋‹จ๊ณ„ | ๋‚ด์šฉ | ์ฃผ์š” ํฌ์ธํŠธ | | | | | | (i) ํ๋ฆ„ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ |

Physics
No Image

Popularity Feedback Constrains Innovation in Cultural Markets

1. ์—ฐ๊ตฌ ์„ค๊ณ„ ๋ฐ ๋ฐฉ๋ฒ•๋ก  | ์š”์†Œ | ๋‚ด์šฉ | | | | | ์‹คํ—˜ ๊ตฌ์กฐ | 128๊ฐœ์˜ ๋…๋ฆฝ์ ์ธ โ€˜์‹œ์žฅโ€™(์‹œ๋“œ ์ด๋ฏธ์ง€ 1๊ฐœ) โ†’ 60์„ธ๋Œ€(์ด 7 680 ์ด๋ฏธ์ง€) ์ง„ํ–‰. ๊ฐ ์„ธ๋Œ€๋งˆ๋‹ค ์ตœ๋Œ€ 12๊ฐœ์˜ ์ด๋ฏธ์ง€๊ฐ€ ์ œ์‹œ๋˜๊ณ , ์ฐธ๊ฐ€์ž๋Š” ํ•˜๋‚˜๋ฅผ ์„ ํƒ ํ›„ 1~24ํ”ฝ์…€์„ ์ˆ˜์ •ํ•ด ์ƒˆ๋กœ์šด ์ด๋ฏธ์ง€๋ฅผ ์ œ์ถœ. | | ์กฐ๊ฑด | PI (Popularity Information) : ์„ ํƒ ๋‹จ๊ณ„์—์„œ ๊ฐ ์ด๋ฏธ์ง€๊ฐ€ ๋ช‡ ๋ฒˆ ์„ ํƒยทํŽธ์ง‘๋๋Š”์ง€ ํ‘œ์‹œ.<br> NPI (No Popularity Information) : ์ธ๊ธฐ ์ •๋ณด ๋น„๊ณต๊ฐœ. | | ์ธก์ • ์ง€ํ‘œ | ๋ฌธํ™” ๋‹ค์–‘์„ฑ :

Computer Science Social Networks
No Image

Population-scale Ancestral Recombination Graphs with tskit 1.0

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ARG์˜ ์ด๋ก ์  ์ค‘์š”์„ฑ vs ์‹ค์šฉ์  ํ•œ๊ณ„ : ์ˆ˜์‹ญ ๋…„๊ฐ„ ์ด๋ก ์  ๋ชจ๋ธ๋กœ๋งŒ ์กด์žฌํ–ˆ์œผ๋‚˜, ๋Œ€๊ทœ๋ชจ ์ƒ˜ํ”Œ์— ๋Œ€ํ•œ ์ถ”๋ก ยท์‹œ๋ฎฌ๋ ˆ์ด์…˜์ด ๋ถˆ๊ฐ€๋Šฅํ–ˆ์Œ. ํ‘œ์ค€ ํฌ๋งทยท์ธํ”„๋ผ ๋ถ€์žฌ : ์„œ๋กœ ๋‹ค๋ฅธ ๋„๊ตฌ๊ฐ€ ์„œ๋กœ ๋‹ค๋ฅธ ๋‚ด๋ถ€ ํ‘œํ˜„์„ ์‚ฌ์šฉํ•ด ๋ฐ์ดํ„ฐ ๊ตํ™˜์ด ์–ด๋ ค์› ์Œ. 2. tskit์˜ ํ•ต์‹ฌ ๊ธฐ์ˆ  | ์š”์†Œ | ์„ค๋ช… | ์˜์˜ | | | | | | Tree Sequence ๋ฐ์ดํ„ฐ ๋ชจ๋ธ | ๋…ธ๋“œ(์‹œ๊ฐ„๋ณ„ ์œ ์ „์ฒด)์™€ ์—ฃ์ง€(์œ ์ „ ๊ตฌ๊ฐ„)๋กœ ๊ตฌ์„ฑ๋œ ํ…Œ์ด๋ธ” ํ˜•ํƒœ | ์†์‹ค ์—†๋Š” ARG ํ‘œํ˜„, ๋ฉ”๋ชจ๋ฆฌยท๋””์Šคํฌ ํšจ์œจ ๊ทน๋Œ€ํ™” | | ๋ฉ”ํƒ€๋ฐ์ดํ„ฐยทํ”„๋กœ๋ฒ ๋„Œ์Šค | ์‚ฌ์ดํŠธ,

Quantitative Biology
No Image

Positional Encoding meets Persistent Homology on Graphs

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ GNN์˜ ํ•œ๊ณ„ : ์ „ํ†ต์ ์ธ ๋ฉ”์‹œ์ง€ ํŒจ์‹ฑ์€ ๋…ธ๋“œ์˜ ๊ตญ์†Œ ์ด์›ƒ ์ •๋ณด์—๋งŒ ์˜์กดํ•ด, ๊ทธ๋ž˜ํ”„ ์ „์ฒด์˜ ์ „์—ญ์  ์œ„์ƒ(์˜ˆ: ์‚ฌ์ดํด, ๊ตฌ๋ฉ) ์ •๋ณด๋ฅผ ๋†“์น˜๊ธฐ ์‰ฝ๋‹ค. ์œ„์น˜ ์ธ์ฝ”๋”ฉ(PE) : Laplacian eigenvectors, random walk ๊ธฐ๋ฐ˜ ๋“ฑ์œผ๋กœ ๋…ธ๋“œ์˜ ์ ˆ๋Œ€ยท์ƒ๋Œ€ ์œ„์น˜๋ฅผ ํŠน์ง•ํ™”ํ•ด GNN์— ์œ„์น˜ ์ •๋ณด๋ฅผ ์ฃผ์ž…ํ•œ๋‹ค. ํ•˜์ง€๋งŒ PE๋Š” ์„ ํƒ๋œ ์ขŒํ‘œ๊ณ„์— ๋ฏผ๊ฐํ•˜๊ณ , ๊ณ ์ฐจ ์œ„์ƒ ๊ตฌ์กฐ๋ฅผ ํฌ์ฐฉํ•˜๊ธฐ ์–ด๋ ต๋‹ค. ์˜์† ๋™ํ˜•๋ก (PH) : ํ•„ํ„ฐ๋ง ํŒŒ๋ผ๋ฏธํ„ฐ๋ฅผ ๋ณ€ํ™”์‹œํ‚ค๋ฉฐ ์ƒ์„ฑ๋˜๋Š” ๋ฐ”์ฝ”๋“œ ๋ฅผ ํ†ตํ•ด ๊ทธ๋ž˜ํ”„์˜ ๋‹ค์ค‘ ์Šค์ผ€์ผ ์œ„์ƒ(0โ€‘์ฐจ, 1โ€‘์ฐจ,

No Image

Predictive Subsampling for Scalable Inference in Networks

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๋Œ€๊ทœ๋ชจ ๋„คํŠธ์›Œํฌ์˜ ๋ณ‘๋ชฉ : ASE์™€ ๊ฐ™์€ ์ŠคํŽ™ํŠธ๋Ÿด ๋ฐฉ๋ฒ•์€ O(nยณ) ๋ณต์žก๋„๋กœ, ์ •์  ์ˆ˜๊ฐ€ ์ˆ˜๋ฐฑ๋งŒ ์ˆ˜์ค€์ด๋ฉด ๋ฉ”๋ชจ๋ฆฌยท์‹œ๊ฐ„ ๋ชจ๋‘ ๋น„ํ˜„์‹ค์ ์ด๋‹ค. ํŠนํžˆ ๋‘ ํ‘œ๋ณธ ๊ฒ€์ •์€ ๋ถ€ํŠธ์ŠคํŠธ๋žฉ ๋ฐ˜๋ณต์ด ํ•„์š”ํ•ด ๋น„์šฉ์ด ๊ธฐํ•˜๊ธ‰์ˆ˜์ ์œผ๋กœ ์ฆ๊ฐ€ํ•œ๋‹ค. ๊ธฐ์กด ์Šค์ผ€์ผ๋ง ์ „๋žต์˜ ํ•œ๊ณ„ : ์ „ํ†ต์ ์ธ ์„œ๋ธŒ์ƒ˜ํ”Œ๋งยท์Šค์ผ€์น˜ยทdivideโ€‘andโ€‘conquer๋Š” ๋„คํŠธ์›Œํฌ ๋ชจ๋ธ์˜ ๋น„์‹๋ณ„์„ฑ(nonโ€‘identifiability) (์˜ˆ: GRDPG์˜ O(p,q) ๋ณ€ํ™˜)๊ณผ ์ •์ โ€‘ํŠน์ • ํŒŒ๋ผ๋ฏธํ„ฐ ๋•Œ๋ฌธ์— ์ง์ ‘ ์ ์šฉ์ด ์–ด๋ ต๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด โ€“ Predictive Sub

Statistics Network
No Image

Prescriptive Scaling Reveals the Evolution of Language Model Capabilities

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์Šค์ผ€์ผ๋ง ๋ฒ•์น™์˜ ํ•œ๊ณ„ : ๊ธฐ์กด์˜ ์‚ฌ์ „ ํ•™์Šต ์Šค์ผ€์ผ๋ง ๋ฒ•์น™์€ ํ‰๊ท  ์†์‹คยทํผํ”Œ๋ ‰์‹œํ‹ฐ์™€ ๊ฐ™์€ ์ง€ํ‘œ์— ์ดˆ์ ์„ ๋งž์ถ”์—ˆ์œผ๋ฉฐ, ์‹ค์ œ ๋ฐฐํฌ ํ™˜๊ฒฝ์—์„œ ์ค‘์š”ํ•œ ํฌ์ŠคํŠธโ€‘ํŠธ๋ ˆ์ด๋‹ ๋‹จ๊ณ„(Instructionโ€‘tuning, RLHF ๋“ฑ)๋ฅผ ๋ฐ˜์˜ํ•˜์ง€ ๋ชปํ•œ๋‹ค. ์‹ค๋ฌด์  ์งˆ๋ฌธ : โ€œ์ฃผ์–ด์ง„ FLOPs ์˜ˆ์‚ฐ์œผ๋กœ ์–ด๋А ์ •๋„์˜ ๋‹ค์šด์ŠคํŠธ๋ฆผ ์ •ํ™•๋„๋ฅผ ๊ธฐ๋Œ€ํ•  ์ˆ˜ ์žˆ๋Š”๊ฐ€?โ€๋ผ๋Š” ์งˆ๋ฌธ์€ ๋ชจ๋ธ ์„ค๊ณ„ยท์˜ˆ์‚ฐ ์ฑ…์ •์— ์ง์ ‘์ ์ธ ๊ฐ€์น˜๋ฅผ ์ œ๊ณตํ•œ๋‹ค. 2. ๋ฐ์ดํ„ฐ์…‹ ๋ฐ ๊ด€์ธก ์„ค๊ณ„ | ๋ฐ์ดํ„ฐ ์ถœ์ฒ˜ | ๊ทœ๋ชจ | ํŠน์ง• | | | | | | Open LLM Leaderboar

Computer Science Machine Learning Model
No Image

Profiling Concurrent Vision Inference Workloads on NVIDIA Jetson -- Extended

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์—ฃ์ง€ AI ๋Š” ๋ฐ์ดํ„ฐ ํ”„๋ผ์ด๋ฒ„์‹œ, ์‹ค์‹œ๊ฐ„ ์‘๋‹ต์„ฑ, ๋„คํŠธ์›Œํฌ ๋น„์šฉ ์ ˆ๊ฐ ๋“ฑ ์—ฌ๋Ÿฌ ์žฅ์ ์„ ์ œ๊ณตํ•˜์ง€๋งŒ, ์ œํ•œ๋œ ์ „๋ ฅยท์—ด ์„ค๊ณ„ ์™€ ๋‹ค์ค‘ ์ž‘์—… ๋™์‹œ ์‹คํ–‰ ์ด๋ผ๋Š” ๋ณตํ•ฉ์ ์ธ ์ œ์•ฝ์„ ๊ฐ€์ง„๋‹ค. Jetson ์‹œ๋ฆฌ์ฆˆ๋Š” GPU + CPU + NPU ๊ฐ€ ํ†ตํ•ฉ๋œ SoC ํ˜•ํƒœ๋กœ, ๋™์‹œ ๋น„์ „ ์ถ”๋ก  ์‹œ ์ž์› ๊ฒฝ์Ÿ์ด ์‹ฌํ™”๋  ๊ฐ€๋Šฅ์„ฑ์ด ํฌ๋‹ค. ๊ธฐ์กด ์—ฐ๊ตฌ๋Š” ์ฃผ๋กœ ๋‹จ์ผ ์›Œํฌ๋กœ๋“œ ์— ์ดˆ์ ์„ ๋งž์ถ”์—ˆ์œผ๋ฉฐ, ๋ฉ€ํ‹ฐโ€‘์ธํผ๋Ÿฐ์Šค ์ƒํ™ฉ์—์„œ์˜ ์„ธ๋ฐ€ํ•œ ์ž์› ์‚ฌ์šฉ ํŒจํ„ด์€ ์ถฉ๋ถ„ํžˆ ํƒ๊ตฌ๋˜์ง€ ์•Š์•˜๋‹ค. 2. ์‹คํ—˜ ์„ค๊ณ„ ๋ฐ ํ”„๋กœํŒŒ์ผ๋ง ๋ฐฉ๋ฒ• | ๋‹จ๊ณ„ | ๋‚ด์šฉ | ์‚ฌ์šฉ ํˆด |

No Image

Proof of Concept: Local TX Real-Time Phase Calibration in MIMO Systems

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์œ„์ƒ ์žก์Œ(Phase Noise) ์€ MIMO ๋ฐ์ดํ„ฐ ์ „์†ก๋Ÿ‰์„ ์ €ํ•˜์‹œํ‚ค๊ณ , ํŠนํžˆ Joint Communication and Sensing (JCnS) ํ™˜๊ฒฝ์—์„œ ํŒŒ๋ผ๋ฏธํ„ฐ ์ถ”์ • ์ •ํ™•๋„๋ฅผ ํฌ๊ฒŒ ๊ฐ์†Œ์‹œํ‚จ๋‹ค. ๊ธฐ์กด 3GPP PTโ€‘RS์™€ ๊ฐ™์€ ์ˆ˜์‹  ์ธก ์œ„์ƒ ๋ณด์ • ๊ธฐ๋ฒ•์€ ์ „์†ก ์ธก ์œ„์ƒ ์ผ๊ด€์„ฑ ์„ ๋ณด์žฅํ•˜์ง€ ๋ชปํ•œ๋‹ค. OTA(Overโ€‘theโ€‘Air) ๊ธฐ๋ฐ˜ ์ƒํ˜ธ๋ณด์ •์€ ๊ณต์œ  ์œ„์ƒ ๊ธฐ์ค€์ด ์—†๊ฑฐ๋‚˜, ์†กยท์ˆ˜์‹  ์žฅ๋น„๊ฐ€ ๋ฌผ๋ฆฌ์ ์œผ๋กœ ๊ฒฐํ•ฉ๋ผ์•ผ ํ•˜๋Š” ์ œ์•ฝ์ด ์žˆ๋‹ค. ๋”ฐ๋ผ์„œ ์†ก์‹  ์ฒด์ธ ๊ฐ„ ๋กœ์ปฌ ์œ„์ƒ ๋ณด์ • ์ด ํ•„์š”ํ•˜๋ฉฐ, ์ด๋Š” ํŠนํžˆ ์ด๋™ํ˜• ๋˜๋Š” ๋ถ„

System Electrical Engineering and Systems Science
No Image

Properties of biodiversity indices that model future extinction risk

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์ƒ๋ฌผ๋‹ค์–‘์„ฑ ๋ณด์ „ ์—์„œ ์ „ํ†ต์ ์œผ๋กœ ์‚ฌ์šฉ๋˜๋Š” Phylogenetic Diversity(PD) ๋Š” ์ข… ํ’๋ถ€๋„๋ณด๋‹ค ์ง„ํ™”์  ์ •๋ณด๋ฅผ ๋” ์ž˜ ๋ฐ˜์˜ํ•œ๋‹ค๋Š” ์ ์—์„œ ์ค‘์š”ํ•˜๊ฒŒ ๋‹ค๋ฃจ์–ด์ ธ ์™”๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ PD ๊ธฐ๋ฐ˜ ์ง€ํ‘œ ๋Š” (โ‘ ) ํŠธ๋ฆฌ ๊ตฌ์กฐ์—๋งŒ ๊ตญํ•œ ๋˜๊ณ , (โ‘ก) ํŠน์ • ๋ฉธ์ข… ์œ„ํ—˜(ฮต) ์„ ๊ณ ๋ คํ•˜์ง€ ๋ชปํ•œ๋‹ค๋Š” ํ•œ๊ณ„๊ฐ€ ์žˆ๋‹ค. EDGE2 ํ”„๋กœ๊ทธ๋žจ์€ ์ด๋Ÿฌํ•œ ํ•œ๊ณ„๋ฅผ ๋ณด์™„ํ•˜๋ ค HEDยทHEDGE๋ฅผ ๋„์ž…ํ–ˆ์ง€๋งŒ, ์—ฌ์ „ํžˆ ๋‹ค์–‘์„ฑ ์ธก์ • ํ•จ์ˆ˜ ฯ†๊ฐ€ ํŠธ๋ฆฌ ํ˜•ํƒœ์—๋งŒ ์ ์šฉ ๋œ๋‹ค๋Š” ์ œ์•ฝ์ด ์กด์žฌํ•œ๋‹ค. 2. ์ฃผ์š” ๊ธฐ์—ฌ | ๊ตฌ๋ถ„ | ๊ธฐ์กด ์ ‘๊ทผ | ๋ณธ ๋…ผ๋ฌธ์˜ ํ™•์žฅ |

Quantitative Biology Model
No Image

Protein Circuit Tracing via Cross-layer Transcoders

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ pLM์˜ ์„ฑ๊ณต : ESM2, AlphaFold ๋“ฑ์€ ์„œ์—ดโ€‘๊ตฌ์กฐโ€‘๊ธฐ๋Šฅ ์˜ˆ์ธก์—์„œ ๋›ฐ์–ด๋‚œ ์„ฑ๋Šฅ์„ ๋ณด์ด๋ฉฐ, ๋‚ด๋ถ€์— ๋ณต์žกํ•œ ๊ตฌ์กฐยท๊ธฐ๋Šฅ ๋ชจํ‹ฐํ”„๋ฅผ ์•”๋ฌต์ ์œผ๋กœ ํ•™์Šตํ•˜๊ณ  ์žˆ์Œ. ํ•ด์„์˜ ํ•œ๊ณ„ : ๊ธฐ์กด SAEs๋Š” ์ธต๋ณ„ ํ‘œํ˜„์„ ํฌ์†Œํ•˜๊ฒŒ ๋ถ„ํ•ดํ•˜์ง€๋งŒ, ์ธตโ€‘๊ฐ„ ์—ฐ์‚ฐ ํ๋ฆ„(โ€œํšŒ๋กœโ€)์„ ํฌ์ฐฉํ•˜์ง€ ๋ชปํ•ด ๋ชจ๋ธ ์ „์ฒด๋ฅผ ์ดํ•ดํ•˜๊ธฐ์— ๋ถ€์กฑํ•จ. ์ „ํ†ต์  ํŠธ๋žœ์Šค์ฝ”๋” : NLP ๋ถ„์•ผ์—์„œ ์ œ์•ˆ๋œ ์ธต๋ณ„ ํŠธ๋žœ์Šค์ฝ”๋”(PLT)๋Š” ๊ฐ MLP๋ฅผ ๋…๋ฆฝ์ ์œผ๋กœ ๊ทผ์‚ฌํ•˜์ง€๋งŒ, ๊นŠ์ด ์ฆ๊ฐ€์— ๋”ฐ๋ผ ๋ˆ„์  ์˜ค๋ฅ˜๊ฐ€ ๋ฐœ์ƒํ•จ. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด โ€“ ๊ต์ฐจโ€‘์ธต ํŠธ๋žœ์Šค์ฝ”๋” (CLT) | ์š”

Computer Science Machine Learning
No Image

Quantifying and Attributing Submodel Uncertainty in Stochastic Simulation Models and Digital Twins

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์„œ๋ธŒ๋ชจ๋ธ ์‚ฌ์šฉ์˜ ํ˜„์‹ค์„ฑ : ์‹ค์ œ ์‹œ์Šคํ…œ์˜ ์ž…๋ ฅยท๊ฒฐ์ •ยท์ตœ์ ํ™” ๋ชจ๋“ˆ์€ ๋ฐ์ดํ„ฐ ๋ถ€์กฑ, ํ”„๋ผ์ด๋ฒ„์‹œ, ์—ฐ์‚ฐ ๋น„์šฉ ๋“ฑ์œผ๋กœ ์ง์ ‘ ์‚ฝ์ž…์ด ์–ด๋ ค์›Œ ๊ทผ์‚ฌ ๋ชจ๋ธ(ํ†ต๊ณ„์ , ๋จธ์‹ ๋Ÿฌ๋‹ ๊ธฐ๋ฐ˜)์œผ๋กœ ๋Œ€์ฒด๋˜๋Š” ๊ฒฝ์šฐ๊ฐ€ ๋นˆ๋ฒˆํ•จ. ๊ธฐ์กด ์—ฐ๊ตฌ์™€์˜ ์ฐจ๋ณ„์  : ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๋ถ„์•ผ์—์„œ๋Š” ์ž…๋ ฅ ๋ถˆํ™•์‹ค์„ฑ(Input Uncertainty, IU) ์— ๋Œ€ํ•œ ์—ฐ๊ตฌ๊ฐ€ ํ™œ๋ฐœํ–ˆ์ง€๋งŒ, ์˜์‚ฌ๊ฒฐ์ •ยท์ œ์–ด ๋กœ์ง ๋“ฑ ๋‚ด๋ถ€ ์„œ๋ธŒ๋ชจ๋ธ์—์„œ ๋ฐœ์ƒํ•˜๋Š” ๋ถˆํ™•์‹ค์„ฑ์€ ๊ฑฐ์˜ ๋‹ค๋ฃจ์ง€ ์•Š์•˜๋‹ค. ๋ณธ ๋…ผ๋ฌธ์€ ์ด๋ฅผ โ€˜์„œ๋ธŒ๋ชจ๋ธ ๋ถˆํ™•์‹ค์„ฑ(Submodel Uncertainty)โ€™ ์œผ๋กœ ์ผ๋ฐ˜ํ™”ํ•˜๊ณ , ํ†ตํ•ฉ์ ์ธ

Statistics Model
No Image

Quantitative enstrophy bounds for measure vorticities

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ Delort ํด๋ž˜์Šค์™€ ์™€๋ฅ˜ ์ง‘์ค‘ : 2์ฐจ์› ์œ ํด๋ฆฌ๋“œ ํ‰๋ฉด ํ˜น์€ ํ† ๋Ÿฌ์Šค (mathbb T^{2}) ์œ„์—์„œ ์ธก์ •๊ฐ’ ์ดˆ๊ธฐ ์™€๋ฅ˜๋ฅผ ํ—ˆ์šฉํ•˜๋Š” ๊ฒฝ์šฐ, Delort(1991)์˜ ๊ฒฐ๊ณผ๊ฐ€ ํ˜„์žฌ๊นŒ์ง€ ๊ฐ€์žฅ ์ผ๋ฐ˜์ ์ธ ์ „์—ญ ์กด์žฌ ๊ฒฐ๊ณผ์ด๋‹ค. ์—ฌ๊ธฐ์„œ๋Š” ์™€๋ฅ˜์˜ โ€œ๋น„์ง‘์ค‘โ€(nonโ€‘concentration) ์กฐ๊ฑด, ์ฆ‰ (M {omega}(r)to0) as (rto0) ๊ฐ€ ํ•ต์‹ฌ์ ์ธ ์—ญํ• ์„ ํ•œ๋‹ค. ์ด์ƒ์†Œ์‚ฐ(anomalous dissipation)๊ณผ Onsager ๊ฐ€์„ค : ์ตœ๊ทผ ์—ฐ๊ตฌ๋Š” ์™€๋ฅ˜ ๋น„์ง‘์ค‘์ด โ€œ์ด์ƒ์†Œ์‚ฐ ์—†์Œโ€(no an

Mathematics
No Image

Quantum Approaches for Dysphonia Assessment in Small Speech Datasets

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๋””์Šคํฌ๋‹ˆ์•„ ์ง„๋‹จ์˜ ์ž„์ƒ์  ์ค‘์š”์„ฑ : ์กฐ๊ธฐ ์ง„๋‹จ์ด ์น˜๋ฃŒ ์„ฑ๊ณต๋ฅ ์„ ํฌ๊ฒŒ ์ขŒ์šฐํ•˜์ง€๋งŒ, ๊ฐ๊ด€์ ์ธ ์Œ์„ฑ ๋ถ„์„ ๋„๊ตฌ๊ฐ€ ๋ถ€์กฑํ•˜๋‹ค. ๋ฐ์ดํ„ฐ ๋ถ€์กฑ ๋ฌธ์ œ : ์˜๋ฃŒ ํ˜„์žฅ์—์„œ ์ˆ˜์ง‘ ๊ฐ€๋Šฅํ•œ ์Œ์„ฑ ๋ฐ์ดํ„ฐ๋Š” ํ™˜์ž ํ”„๋ผ์ด๋ฒ„์‹œ, ๋น„์šฉ, ์žฅ๋น„ ์ œ์•ฝ ๋“ฑ์œผ๋กœ ์ œํ•œ์ ์ด๋‹ค. ์ด๋Š” ๋”ฅ๋Ÿฌ๋‹ ๋ชจ๋ธ์ด ๊ณผ์ ํ•ฉ(overfitting)๋˜๊ธฐ ์‰ฌ์šด ํ™˜๊ฒฝ์„ ๋งŒ๋“ ๋‹ค. 2. ๋ฐฉ๋ฒ•๋ก  | ๋‹จ๊ณ„ | ๋‚ด์šฉ | ๋น„๊ณ  | | | | | | ๋ฐ์ดํ„ฐ ์ „์ฒ˜๋ฆฌ | ์›์‹œ ์Œ์„ฑ์„ 128โ€‘ํ”„๋ ˆ์ž„ ๋ฉœ ์ŠคํŽ™ํŠธ๋กœ๊ทธ๋žจ์œผ๋กœ ๋ณ€ํ™˜ | ์‹œ๊ฐ„โ€‘์ฃผํŒŒ์ˆ˜ ์ •๋ณด๋ฅผ ๋ณด์กด, CNN/QNN ์ž…๋ ฅ ํ˜•ํƒœ์— ์ ํ•ฉ | |

Data
No Image

RainyGS: Efficient Rain Synthesis with Physically-Based Gaussian Splatting

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ NeRFยท3DGS ํ•œ๊ณ„ : ๊ธฐ์กด ์‹ ๊ฒฝ ์žฅ๋ฉด ํ‘œํ˜„์€ ์ƒˆ๋กœ์šด ์‹œ์  ํ•ฉ์„ฑ์— ๊ฐ•ํ•˜์ง€๋งŒ, ๋ฌผ๋ฆฌ ๊ธฐ๋ฐ˜ ๋™์  ํ˜„์ƒ(๋น„, ๋ˆˆ, ์—ฐ๊ธฐ ๋“ฑ) ์‚ฝ์ž…์—๋Š” ๋ฌผ๋ฆฌ ๋ฒ•์น™์„ ์ง์ ‘ ๋ฐ˜์˜ํ•˜๊ธฐ ์–ด๋ ค์›€. ์ „ํ†ต ์‹œ๋ฎฌ๋ ˆ์ด์…˜์˜ ๋ฌธ์ œ์  : ๊ณ ํ’ˆ์งˆ ๋ฌผ๋ฆฌ ์‹œ๋ฎฌ๋ ˆ์ด์…˜์€ ์ •ํ™•ํ•˜์ง€๋งŒ, ์”ฌ ๊ตฌ์ถ•ยท๋ Œ๋”๋ง ํŒŒ์ดํ”„๋ผ์ธ์ด ๋ณต์žกํ•˜๊ณ  ์‹ค์‹œ๊ฐ„ ์ ์šฉ์ด ์–ด๋ ค์›Œ ๋Œ€๊ทœ๋ชจ ์˜คํ”ˆ์›”๋“œ์— ์ ์šฉํ•˜๊ธฐ ํž˜๋“ฆ. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด ๋ฌผ๋ฆฌ ๊ธฐ๋ฐ˜ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ + 3DGS : ๋น—๋ฐฉ์šธยท๋ฌผ๋ณด๋ผ์˜ ๋ฌผ๋ฆฌ ๋ชจ๋ธ(์ค‘๋ ฅ, ๊ณต๊ธฐ ์ €ํ•ญ, ํ‘œ๋ฉด ์žฅ๋ ฅ ๋“ฑ)์„ ๊ฐ€์šฐ์‹œ์•ˆ ์Šคํ”Œ๋ž˜ํŒ… ํ”„๋ ˆ์ž„์›Œํฌ์— ๋งคํ•‘. ๊ฐ€์šฐ์‹œ์•ˆ ์Šคํ”Œ๋ž˜ํŒ…์˜ ์žฅ์ 

No Image

Random Wavelet Features for Graph Kernel Machines

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ๊ทธ๋ž˜ํ”„ ์ปค๋„ ์€ ์ •์  ๊ฐ„ ์œ ์‚ฌ์„ฑ์„ ์ •์˜ํ•˜๋Š” ๊ฐ•๋ ฅํ•œ ๋„๊ตฌ์ด์ง€๋งŒ, ์ปค๋„ ํ–‰๋ ฌ์„ ์ง์ ‘ ๊ณ„์‚ฐํ•˜๋ฉด ๋ฉ”๋ชจ๋ฆฌยท์‹œ๊ฐ„ ๋ณต์žก๋„๊ฐ€ (O(N^2)) ~ (O(N^3)) ์œผ๋กœ ๊ธ‰์ฆํ•œ๋‹ค. ๋žœ๋ค ํ”ผ์ฒ˜(RFF) ๋Š” ์œ ํด๋ฆฌ๋“œ ๊ณต๊ฐ„์—์„œ ์ปค๋„์„ ์ €์ฐจ์› ๋‚ด์  ํ˜•ํƒœ๋กœ ๊ทผ์‚ฌํ•ด ํ•™์Šต ํšจ์œจ์„ฑ์„ ํฌ๊ฒŒ ๋†’์˜€๋‹ค. ๊ทธ๋ž˜ํ”„ ๊ตฌ์กฐ์— ๋งž๋Š” ์ŠคํŽ™ํŠธ๋Ÿผ ๊ธฐ๋ฐ˜ ๋žœ๋ค ํ”ผ์ฒ˜ ๊ฐ€ ์•„์ง ์ถฉ๋ถ„ํžˆ ์—ฐ๊ตฌ๋˜์ง€ ์•Š์•˜์œผ๋ฉฐ, ํŠนํžˆ ์ŠคํŽ™ํŠธ๋Ÿผ์ด ์ข๊ฒŒ ์ง‘์ค‘๋œ(์ŠคํŽ™ํŠธ๋Ÿผ ๋กœ์ปฌ๋ผ์ด์ฆˆ๋“œ) ์ปค๋„ ์— ๋Œ€ํ•œ ํšจ์œจ์ ์ธ ๊ทผ์‚ฌ ๋ฐฉ๋ฒ•์ด ํ•„์š”ํ–ˆ๋‹ค. 2. ํ•ต์‹ฌ ์•„์ด๋””์–ด 1. ๊ทธ๋ž˜ํ”„ ์‹ ํ˜ธ ์ฒ˜๋ฆฌ(GSP)์™€ ์›จ์ด๋ธŒ

Computer Science Machine Learning
No Image

Randomized Zero Forcing

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ์ œ๋กœ ํฌ์‹ฑ(zero forcing) ์€ ์›๋ž˜ deterministic ์ƒ‰๋ณ€๊ฒฝ ๊ทœ์น™์œผ๋กœ, ๊ทธ๋ž˜ํ”„ ์ด๋ก ยทํ–‰๋ ฌ ์ด๋ก (์ตœ์†Œ ๋žญํฌ) ๋ฐ ์–‘์ž ์ œ์–ด์— ํ™œ์šฉ๋˜์—ˆ๋‹ค. ํ™•๋ฅ ์  ์ œ๋กœ ํฌ์‹ฑ(PZF) ๋Š” ๋ฌด๋ฐฉํ–ฅ ๊ทธ๋ž˜ํ”„์—์„œ ๊ฐ ํŒŒ๋ž€ ์ •์ ์ด ์ธ์ ‘ํ•œ ํฐ ์ •์ ์„ ๋…๋ฆฝ์ ์œผ๋กœ ์ผ์ • ํ™•๋ฅ ๋กœ ์ƒ‰์„ ๋ฐ”๊พธ๊ฒŒ ํ•˜๋Š” ๋ชจ๋ธ์ด๋ฉฐ, ์ „ํŒŒ๊ฐ€ ํ•ญ์ƒ ์ „์ฒด ๊ทธ๋ž˜ํ”„์— ๋„๋‹ฌํ•œ๋‹ค๋Š” ์ ์ด ํŠน์ง•์ด๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ์‹ค์ œ ์‹œ์Šคํ…œ (์˜ˆ: ๊ณต๊ธ‰๋ง, ๊ธˆ์œต ๋„คํŠธ์›Œํฌ)์€ ๋ฐฉํ–ฅ์„ฑ ๊ณผ ๊ฐ€์ค‘์น˜(์˜์กด๋„) ๊ฐ€ ํ•ต์‹ฌ์ด๋‹ค. ์œ„ํ—˜์€ ๋ฐ›๋Š” ์ชฝ ์˜ ์˜์กด ๋น„์œจ์— ๋”ฐ๋ผ ์ „ํŒŒ๋˜๋ฏ€๋กœ, PZF์˜ ๋ฌด๋ฐฉํ–ฅยท๋™๋“ฑ ํ™•๋ฅ 

Mathematics
No Image

Rate-Distortion Optimization for Ensembles of Non-Reference Metrics

| ํ•ญ๋ชฉ | ๋‚ด์šฉ ๋ฐ ํ‰๊ฐ€ | | | | | ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ | โ€ข UGC๋Š” ์›๋ณธ์ด ์žก์Œยท๋ธ”๋Ÿฌยท์••์ถ• ์•„ํ‹ฐํŒฉํŠธ ๋“ฑ์„ ํฌํ•จํ•ด ์ „์ฐธ์กฐ ์ง€ํ‘œ๊ฐ€ ์‹ค์ œ ํ’ˆ์งˆ์„ ๊ณผ๋Œ€ํ‰๊ฐ€ํ•œ๋‹ค.<br>โ€ข ์ตœ์‹  NRM(์˜ˆ: QualiCLIP, TOPIQโ€‘NR ๋“ฑ)์€ ๋”ฅ๋Ÿฌ๋‹ ๊ธฐ๋ฐ˜์œผ๋กœ ๋†’์€ ์ƒ๊ด€๊ด€๊ณ„๋ฅผ ๋ณด์ด์ง€๋งŒ, RDO์— ์ง์ ‘ ์ ์šฉํ•˜๋ฉด ์—ฐ์‚ฐ๋Ÿ‰์ด ๊ธ‰์ฆํ•œ๋‹ค. | | ๊ธฐ์กด LNRM ํ•œ๊ณ„ | โ€ข NRM์˜ ๊ทธ๋ž˜๋””์–ธํŠธ๊ฐ€ โ€œlocally unstableโ€ํ•ด ์ž‘์€ ์ž…๋ ฅ ๋ณ€ํ™”์—๋„ ๊ธ‰๊ฒฉํžˆ ๋ณ€ํ•จ.<br>โ€ข ๋‹จ์ผ NRM์— ์ตœ์ ํ™”ํ•˜๋ฉด ํ•ด๋‹น NRM์—์„œ๋Š” BDโ€‘rate ์ ˆ๊ฐ์ด ํฌ์ง€๋งŒ, ๋‹ค๋ฅธ

Image Processing Electrical Engineering and Systems Science
No Image

Reactive Slip Control in Multifingered Grasping: Hybrid Tactile Sensing and Internal-Force Optimization

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ๋‹ค์ค‘์†๊ฐ€๋ฝ ๊ทธ๋ฆฌํผ ๋Š” ๋ณต์žกํ•œ ๋ฌผ์ฒด ์กฐ์ž‘์— ํ•„์ˆ˜์ ์ด์ง€๋งŒ, ๊ฐ ์†๊ฐ€๋ฝ์— ๊ฐ€ํ•ด์ง€๋Š” ํž˜์„ ์กฐํ™”๋กญ๊ฒŒ ๋ฐฐ๋ถ„ํ•˜์ง€ ์•Š์œผ๋ฉด ๋ฌผ์ฒด ๋ ˆ๋ฒจ ์™€ํฌ ๊ฐ€ ๋ณ€ํ•ด ์ž์„ธ๊ฐ€ ๋ถˆ์•ˆ์ •ํ•ด์ง„๋‹ค. ๊ธฐ์กด์˜ ๋ฐ˜์‘ํ˜• ์Šฌ๋ฆฝ ์ œ์–ด(RSC) ๋Š” ๋ชจ๋“  ์†๊ฐ€๋ฝ์— ๋™์ผํ•œ ํž˜์„ ์ฆ๊ฐ€์‹œํ‚ค๋Š” ๋‹จ์ˆœ ๋ฐฉ์‹์„ ์‚ฌ์šฉํ•ด, ๋‹ค์ค‘์†๊ฐ€๋ฝ ์‹œ์Šคํ…œ์—์„  ๋ถ€์ž‘์šฉ์ด ํฌ๋‹ค. ๋งˆ์ฐฐ๊ณ„์ˆ˜ยท์™ธ๋ถ€ ํ•˜์ค‘ ๋“ฑ ์ •ํ™•ํ•œ ํŒŒ๋ผ๋ฏธํ„ฐ๋ฅผ ์‚ฌ์ „์— ์•Œ๊ธฐ ์–ด๋ ค์šด ์‹ค์ œ ํ™˜๊ฒฝ์—์„œ๋Š” ๋ฐ์ดํ„ฐโ€‘๋“œ๋ฆฌ๋ธ ์ ‘๊ทผ๊ณผ ๋ชจ๋ธโ€‘๋ฒ ์ด์Šค ์ ‘๊ทผ์„ ๊ฒฐํ•ฉํ•œ ํ•˜์ด๋ธŒ๋ฆฌ๋“œ๊ฐ€ ์š”๊ตฌ๋œ๋‹ค. 2. ํ•ต์‹ฌ ๊ธฐ์—ฌ | ๋ฒˆํ˜ธ | ๋‚ด์šฉ | ์˜์˜ | | | | | |โ‘ | ๋‚ด๋ถ€

Robotics Computer Science
No Image

Real time fault detection in 3D printers using Convolutional Neural Networks and acoustic signals

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ 3D ํ”„๋ฆฐํŒ…์˜ ์‚ฐ์—…์  ์ค‘์š”์„ฑ : ํ•ญ๊ณตยท์ž๋™์ฐจ ๋“ฑ ๊ณ ์‹ ๋ขฐ์„ฑ์ด ์š”๊ตฌ๋˜๋Š” ๋ถ„์•ผ์—์„œ ๋ถ€ํ’ˆ ์ƒ์‚ฐ์— ํ™œ์šฉ๋„๊ฐ€ ๊ธ‰์ฆํ•˜๊ณ  ์žˆ๋‹ค. ์ „ํ†ต์  ๊ฒฐํ•จ ํƒ์ง€์˜ ํ•œ๊ณ„ : ์‹œ๊ฐ ๊ฒ€์‚ฌ : ์ธ๊ฐ„์˜ ์ฃผ๊ด€์„ฑ, ์—ฐ์† ๋ชจ๋‹ˆํ„ฐ๋ง ๋ถˆ๊ฐ€, ์ž‘์€ ๊ฒฐํ•จ ํƒ์ง€ ์–ด๋ ค์›€. ์ ‘์ด‰ํ˜• ์„ผ์„œ : ์„ค์น˜ยท์œ ์ง€ ๋น„์šฉ ๋†’์Œ, ํ”„๋ฆฐํ„ฐ ๊ตฌ์กฐ ๋ณ€๊ฒฝ ํ•„์š”, ์ œํ•œ๋œ ๊ฐ์ง€ ๋ฒ”์œ„. ์นด๋ฉ”๋ผ ๊ธฐ๋ฐ˜ : ์‹œ์•ผ ์ œํ•œ, ์กฐ๋ช…ยทํ™˜๊ฒฝ ๋ณ€ํ™”์— ๋ฏผ๊ฐ, ๋‹ค์ค‘ ์นด๋ฉ”๋ผ ์‹œ ๋น„์šฉยท๋ณต์žก๋„ ์ƒ์Šน. ์Œํ–ฅ ๊ธฐ๋ฐ˜ ์ ‘๊ทผ : ๋น„์ ‘์ด‰์‹, ์ €๋น„์šฉ, ํ”„๋ฆฐํ„ฐ ๋‚ด๋ถ€ ๊ตฌ์กฐ๋ฅผ ๋ณ€๊ฒฝํ•˜์ง€ ์•Š์•„๋„ ๋˜๋ฉฐ, ๊ธฐ๊ณ„ ์ž‘๋™ ์ค‘ ๋ฐœ์ƒํ•˜๋Š” ๋ฏธ์„ธํ•œ ์†Œ๋ฆฌ

Network Electrical Engineering and Systems Science Detection
No Image

Real-time Range-Angle Estimation and Tag Localization for Multi-static Backscatter Systems

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ ์•ฐ๋น„์–ธํŠธ IoT์™€ 6G : ์ดˆ์ €์ „๋ ฅ ๋””๋ฐ”์ด์Šค๊ฐ€ ๋Œ€๊ทœ๋ชจ๋กœ ๋ฐฐ์น˜๋˜๋Š” 6G ํ™˜๊ฒฝ์—์„œ๋Š” ์ „๋ ฅ ๊ณต๊ธ‰์ด ์–ด๋ ค์šด ์„ผ์„œยทํƒœ๊ทธ์— ๋ฐฑ์Šค์บํ„ฐ ํ†ต์‹ ์ด ์ตœ์ ์ด๋‹ค. ๊ธฐ์กด ๋‹จ์ผ์ •์ (monostatic) ๋ฐฑ์Šค์บํ„ฐ๋Š” ๋ฒ”์œ„ยทํ™•์žฅ์„ฑ์— ํ•œ๊ณ„๊ฐ€ ์žˆ์—ˆ์ง€๋งŒ, ๋‹ค์ค‘์ •์ ยท์•ฐ๋น„์–ธํŠธ ๊ตฌ์กฐ๋Š” ์ด๋Ÿฌํ•œ ์ œ์•ฝ์„ ๊ทน๋ณตํ•œ๋‹ค. ์œ„์น˜์ถ”์ •์˜ ์ค‘์š”์„ฑ : ์ž์‚ฐ ์ถ”์ , ์ปจํ…์ŠคํŠธ ์ธ์‹ ๋“ฑ ์‹ค์‹œ๊ฐ„ ์œ„์น˜์ •๋ณด๊ฐ€ ํ•„์ˆ˜์ ์ธ ์„œ๋น„์Šค๊ฐ€ ๋Š˜์–ด๋‚˜๋ฉด์„œ, ํ†ต์‹ ๊ณผ ๋™์‹œ์— ์ •ํ™•ํ•œ ์œ„์น˜์ถ”์ •์ด ๊ฐ€๋Šฅํ•œ ์‹œ์Šคํ…œ ์„ค๊ณ„๊ฐ€ ์š”๊ตฌ๋œ๋‹ค. 2. ์‹œ์Šคํ…œ ์•„ํ‚คํ…์ฒ˜ ๋‹ค์ค‘์ •์  ๊ตฌ์„ฑ : ๋‹ค์ˆ˜์˜ ๋‹จ์ผ์•ˆํ…Œ๋‚˜ TX, ๋‹ค์ค‘์•ˆํ…Œ๋‚˜ R

System Electrical Engineering and Systems Science
No Image

RefineFormer3D: Efficient 3D Medical Image Segmentation via Adaptive Multi-Scale Transformer with Cross Attention Fusion

1. ์—ฐ๊ตฌ ๋™๊ธฐ ๋ฐ ๋ฐฐ๊ฒฝ CNNโ€‘๊ธฐ๋ฐ˜ Uโ€‘Net ๊ณ„์—ด ์€ ์ง€์—ญ์  ํŠน์ง•์— ๊ฐ•ํ•˜์ง€๋งŒ ์ „์—ญ ์ปจํ…์ŠคํŠธ ๋ถ€์กฑ์ด 3D ๋ณต์žก ๊ตฌ์กฐ ๋ถ„ํ• ์— ํ•œ๊ณ„๊ฐ€ ๋œ๋‹ค. Transformer ๊ธฐ๋ฐ˜ ๋ชจ๋ธ(TransUNet, UNETR, Swinโ€‘UNet ๋“ฑ)์€ ์ „์—ญ selfโ€‘attention์œผ๋กœ ์„ฑ๋Šฅ์„ ๋Œ์–ด์˜ฌ๋ ธ์ง€๋งŒ, ๋ฉ”๋ชจ๋ฆฌยท์—ฐ์‚ฐ ๋น„์šฉ ์ด ๊ธ‰์ฆํ•ด ์ž„์ƒ ์ ์šฉ์ด ์–ด๋ ค์› ๋‹ค. ๊ธฐ์กด ์Šคํ‚ต ์—ฐ๊ฒฐ์€ ์ •์  concat ํ˜น์€ ๊ณ ๋น„์šฉ ์–ดํ…์…˜ ์— ์˜์กดํ•ด ๋‹ค์ค‘ ์Šค์ผ€์ผ ํŠน์ง•์„ ์„ ํƒ์ ์œผ๋กœ ํ™œ์šฉ ํ•˜์ง€ ๋ชปํ•œ๋‹ค๋Š” ์ ์ด ์ง€์ ๋œ๋‹ค. 2. ํ•ต์‹ฌ ๊ธฐ์—ฌ ๋ฐ ๊ธฐ์ˆ ์  ํ˜์‹  | ๊ตฌ์„ฑ ์š”์†Œ | ๊ธฐ์กด ๋ฐฉ๋ฒ•๊ณผ ์ฐจ๋ณ„

Image Processing Electrical Engineering and Systems Science
No Image

Regret and Sample Complexity of Online Q-Learning via Concentration of Stochastic Approximation with Time-Inhomogeneous Markov Chains

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ์˜จ๋ผ์ธ Qโ€‘๋Ÿฌ๋‹ ์€ ์—์ด์ „ํŠธ๊ฐ€ ํ™˜๊ฒฝ๊ณผ ์ƒํ˜ธ์ž‘์šฉํ•˜๋ฉด์„œ Qโ€‘๊ฐ’์„ ๋™์‹œ์— ์—…๋ฐ์ดํŠธํ•˜๋Š” ์ „ํ˜•์ ์ธ ๋ชจ๋ธโ€‘ํ”„๋ฆฌ ๋ฐฉ๋ฒ•์ด๋‹ค. ๊ธฐ์กด ์ด๋ก ์€ ์ฃผ๋กœ ์˜คํ”„๋ผ์ธ (๊ณ ์ • ์ •์ฑ…์— ์˜ํ•ด ์ƒ์„ฑ๋œ ๋ฐ์ดํ„ฐ) ์ƒํ™ฉ์—์„œ ์ˆ˜๋ ด์„ฑ์„ ๋‹ค๋ฃจ์—ˆ์œผ๋ฉฐ, ํ›„ํšŒ(regret) ๋‚˜ ์ƒ˜ํ”Œ ๋ณต์žก๋„ ์— ๋Œ€ํ•œ ๊ณ ํ™•๋ฅ  ๋ณด์žฅ์€ ๊ฑฐ์˜ ์—†์—ˆ๋‹ค. ์‹ค์ œ ๋”ฅ Qโ€‘๋„คํŠธ์›Œํฌ(DQN)์™€ ๊ฐ™์€ ์ตœ์‹  RL ์‹œ์Šคํ…œ์€ ฮตโ€‘greedy ํ˜น์€ Boltzmann(softmax) ํƒ์ƒ‰์„ ์‚ฌ์šฉํ•˜์ง€๋งŒ, ์ด๋Ÿฌํ•œ ํƒ์ƒ‰ ์ „๋žต์— ๋Œ€ํ•œ ์ด๋ก ์  ๋ณด์žฅ์€ ๋ถ€์žฌํ–ˆ๋‹ค. ๋”ฐ๋ผ์„œ ๋‚™๊ด€์ฃผ์˜ ์—†์ด ์ˆœ์ˆ˜ ๋ชจ๋ธโ€‘ํ”„๋ฆฌ Qโ€‘๋Ÿฌ๋‹ ์ด ์–ผ

Computer Science Learning Machine Learning
No Image

Relativistic nuclear recoil effects in hyperfine splitting of hydrogenic systems

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋™๊ธฐ ํ•ต ๋ฐ˜๋™ ํšจ๊ณผ ๋Š” ์›์ž ์ˆ˜์ค€์—์„œ ์งˆ๋Ÿ‰๋น„ (m/M)๊ฐ€ ์ž‘์€ ๊ฒฝ์šฐ์—๋„ ์ •๋ฐ€ ๊ณ„์‚ฐ์— ํ•„์ˆ˜์ ์ด๋‹ค. ๋น„์ƒ๋Œ€๋ก ์  QED(NRQED) ์™€ ์ค‘์ž…์ž QED(HPQED) ๋Š” ๊ฐ๊ฐ ์ €์—๋„ˆ์ง€( (ksim malpha) )์™€ ๊ณ ์—๋„ˆ์ง€((ksim m)) ์˜์—ญ์„ ํšจ์œจ์ ์œผ๋กœ ๋‹ค๋ฃจ๋Š” ์ฒด๊ณ„์ด๋‹ค. ๋‘ ์ ‘๊ทผ๋ฒ•์„ ๋™์‹œ์— ์‚ฌ์šฉํ•จ์œผ๋กœ์จ ์ „ ์˜์—ญ์„ ํฌ๊ด„ํ•˜๋Š” ๊ณ„์‚ฐ์ด ๊ฐ€๋Šฅํ•ด์กŒ๋‹ค. ๊ธฐ์กด Bodwinโ€‘Yennie ๊ณ„์‚ฐ์€ 30๋…„ ์ „์˜ ๊ฒฐ๊ณผ๋กœ, ์ตœ์‹  ์‹คํ—˜(ํŠนํžˆ 1S ์ดˆ๋ฏธ์„ธ ๊ตฌ์กฐ ๋ถ„ํ• )๊ณผ์˜ ๋ถˆ์ผ์น˜๊ฐ€ ์ง€์†๋˜์–ด ์™”๋‹ค. 2. ์ด๋ก ์  ์ ‘๊ทผ๋ฒ• | ๋‹จ๊ณ„ |

Physics System
No Image

Reply To: Global Gridded Population Datasets Systematically Underrepresent Rural Population by Josias Lรกng-Ritter et al

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ์› ๋…ผ๋ฌธ์˜ ํ•ต์‹ฌ ์ฃผ์žฅ ์› ๋…ผ๋ฌธ์€ 2000โ€‘2010๋…„ ์‚ฌ์ด ์ „ ์„ธ๊ณ„ 307๊ฐœ ๋Œ ์žฌ์ •์ฐฉ ์ง€์—ญ์„ ๋Œ€์ƒ์œผ๋กœ, ์ฃผ์š” ๊ฒฉ์ž ์ธ๊ตฌ ๋ฐ์ดํ„ฐ์…‹(GRUMP, WorldPop, GHSโ€‘POP ๋“ฑ)์˜ ๋†์ดŒ ์ธ๊ตฌ ์ถ”์ •์น˜๊ฐ€ ์‹ค์ œ ์žฌ์ •์ฐฉ ์ธ๊ตฌ๋ณด๋‹ค 80%~โ€‘32% ๋‚ฎ๋‹ค๊ณ  ๊ฒฐ๋ก ์ง“๋Š”๋‹ค. ์ด๋ฅผ ๊ทผ๊ฑฐ๋กœ โ€œ๊ฐ€์žฅ ์ •ํ™•ํ•œ ๋ฐ์ดํ„ฐ์…‹์ด๋ผ ํ• ์ง€๋ผ๋„ ๋†์ดŒ ์ธ๊ตฌ๊ฐ€ ์ ˆ๋ฐ˜ ๊ฐ€๋Ÿ‰ ๊ณผ์†Œ๊ณ„์‚ฐ๋œ๋‹คโ€๋Š” ๊ฐ•๋ ฅํ•œ ์ฃผ์žฅ์„ ์ œ์‹œํ•œ๋‹ค. 2. ํ‘œ๋ณธ ์„ ์ •๊ณผ ์‹œ๊ณ„์—ด ๋ถˆ์ผ์น˜ ์‹œ๊ณต๊ฐ„ ๋ฏธ์Šค๋งค์น˜ : ๋Œ ํ”„๋กœ์ ํŠธ๋Š” ์ฐฉ๊ณตยทํ† ์ง€ ์ˆ˜์šฉ ๋‹จ๊ณ„๋ถ€ํ„ฐ ์ˆ˜์‹ญ ๋…„์— ๊ฑธ์ณ ์ง„ํ–‰๋œ๋‹ค. ์› ๋…ผ๋ฌธ์€ โ€˜์™„๊ณต 10๋…„ ์ „โ€™

Data System Quantitative Biology
No Image

Reproducibility and Statistical Methodology

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ œ๊ธฐ ์žฌํ˜„์„ฑ ์œ„๊ธฐ : Nosek et al. (2015) OSCโ€‘RP๋Š” 100๊ฐœ์˜ ์‹ฌ๋ฆฌํ•™ ๋…ผ๋ฌธ์„ ๋ณต์ œํ–ˆ์„ ๋•Œ, ์› ๋…ผ๋ฌธ์˜ 97 %๊ฐ€ ํ†ต๊ณ„์ ์œผ๋กœ ์œ ์˜ํ–ˆ์Œ์—๋„ ๋ณต์ œ ๊ฒฐ๊ณผ๋Š” 36 %๋งŒ์ด ์œ ์˜ํ–ˆ๋‹ค๋Š” ์ถฉ๊ฒฉ์ ์ธ ์ˆ˜์น˜๋ฅผ ์ œ์‹œํ–ˆ๋‹ค. ๋Œ€์กฐ ์—ฐ๊ตฌ : Etz & Vandekerckhove (2016), Klein et al. (2014) (Many Labs), Camerer et al. (2016) ๋“ฑ์€ ๋™์ผํ•œ ๋ณต์ œ ํ”„๋กœํ† ์ฝœ์—๋„ ๋ถˆ๊ตฌํ•˜๊ณ  60 % ~ 85 %์˜ ๋†’์€ ์žฌํ˜„์„ฑ์„ ๋ณด๊ณ , OSC ๊ฒฐ๊ณผ๊ฐ€ ๊ณผ๋„ํ•˜๊ฒŒ ๋น„๊ด€์ ์ผ ๊ฐ€๋Šฅ์„ฑ์„ ์ œ๊ธฐ

Statistics Applications
No Image

Resp-Agent: An Agent-Based System for Multimodal Respiratory Sound Generation and Disease Diagnosis

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ๋ฌธ์ œ ์ •์˜ | ๋ฌธ์ œ | ๊ธฐ์กด ์ ‘๊ทผ์˜ ํ•œ๊ณ„ | | | | | ํ‘œํ˜„ ๋ณ‘๋ชฉ | ์˜ค๋””์˜ค โ†’ ๋ฉœโ€‘์ŠคํŽ™ํŠธ๋กœ๊ทธ๋žจ ๋ณ€ํ™˜ ์‹œ ์œ„์ƒยท๋ฏธ์„ธ ์‹œ๊ฐ„ ๊ตฌ์กฐ ์†์‹ค, ํ…์ŠคํŠธ ๋ชจ๋ธ์€ ๊ฐ๊ด€์  ์Œํ–ฅ ์ฆ๊ฑฐ ๋ถ€์žฌ | | ๋ฐ์ดํ„ฐ ๋ถ€์กฑยท๋ถˆ๊ท ํ˜• | ๊ณต๊ฐœ ํ˜ธํก์Œ ๋ฐ์ดํ„ฐ์…‹ ๊ทœ๋ชจ ์ž‘๊ณ  ํด๋ž˜์Šค๊ฐ€ ์ œํ•œ์ ์ด๋ฉฐ, ๋ฉ”ํƒ€๋ฐ์ดํ„ฐ ํ™œ์šฉ์ด ๋ฏธํก | 2. ์ฃผ์š” ๊ธฐ์—ฌ | ๋ฒˆํ˜ธ | ๊ธฐ์—ฌ ๋‚ด์šฉ | ์˜์˜ | | | | | | โ‘  Respโ€‘229k | 229 101๊ฐœ์˜ ๊ณ ํ’ˆ์งˆ ํ˜ธํก์Œ + ํ‘œ์ค€ํ™”๋œ ์ž„์ƒ ์š”์•ฝ, 16๊ฐœ ํด๋ž˜์Šค, ์—„๊ฒฉํ•œ OOD ํ…Œ์ŠคํŠธ ์Šคํ”Œ๋ฆฟ ์ œ๊ณต | ๋ฉ€ํ‹ฐ๋ชจ๋‹ฌ ํ•™์Šตยทํ‰๊ฐ€๋ฅผ ์œ„ํ•œ

System Audio Processing Electrical Engineering and Systems Science
No Image

Rethinking Diffusion Models with Symmetries through Canonicalization with Applications to Molecular Graph Generation

| ๊ตฌ๋ถ„ | ๋‚ด์šฉ | | | | | ํ•ต์‹ฌ ์•„์ด๋””์–ด | ๋Œ€์นญ์„ ๊นจ๋œจ๋ฆฌ๊ณ  (canonical pose/order) ์ •๊ทœํ™”๋œ ์Šฌ๋ผ์ด์Šค ์—์„œ ๋น„๋“ฑ๋ณ€ ๋ชจ๋ธ์„ ํ•™์Šตํ•œ ๋’ค, ์ƒ์„ฑ ์‹œ ๋ฌด์ž‘์œ„ ๋Œ€์นญ ๋ณ€ํ™˜์„ ์ ์šฉํ•ด ๋ถˆ๋ณ€ ๋ถ„ํฌ๋ฅผ ๋ณต์›ํ•œ๋‹ค. ์ด๋Š” โ€œquotientโ€‘spaceโ€ ๊ด€์ ์—์„œ ๊ถค๋„ ๋Œ€ํ‘œ ์„ ํƒ ์ด๋ผ๋Š” ์ˆ˜ํ•™์  ์ ˆ์ฐจ์™€ ๋™์ผํ•˜๋‹ค. | | ์ด๋ก ์  ๊ธฐ์—ฌ | 1. Factorization Theorem (์ •๊ทœํ™”๋œ ์Šฌ๋ผ์ด์Šค์™€ Haar ๋ฌด์ž‘์œ„ํ™”๊ฐ€ ๋ชจ๋“  Gโ€‘๋ถˆ๋ณ€ ๋ถ„ํฌ๋ฅผ ์™„์ „ํžˆ ํ‘œํ˜„ํ•œ๋‹ค) <br>2. Universality : ๋น„๋“ฑ๋ณ€ ๋ฐฑ๋ณธ์ด ์ •๊ทœํ™”๋œ ์Šฌ๋ผ์ด์Šค์—์„œ

Computer Science Machine Learning Model
No Image

retinalysis-vascx: An explainable software toolbox for the extraction of retinal vascular biomarkers

1. ์—ฐ๊ตฌ ๋ฐฐ๊ฒฝ ๋ฐ ํ•„์š”์„ฑ | ๋‚ด์šฉ | ์„ค๋ช… | | | | | Oculomics | ์•ˆ๊ตฌ ์˜์ƒ์„ ํ†ตํ•ด ์ „์‹  ์งˆํ™˜(๊ณ ํ˜ˆ์••, ์‹ ์žฅ์งˆํ™˜, ๋‡Œ์กธ์ค‘ ๋“ฑ) ์œ„ํ—˜์„ ์˜ˆ์ธกํ•˜๋Š” ๋ถ„์•ผ. | | ๊ธฐ์กด ํˆด | PVBM, AutoMorph ๋“ฑ์€ ๋ฐ”์ด์˜ค๋งˆ์ปค ๊ณ„์‚ฐ์„ ์ง€์›ํ•˜์ง€๋งŒ ์‹œ์‹ ๊ฒฝ์œ ๋‘โ€‘ํ™ฉ๋ฐ˜ ์ถ• ๊ธฐ๋ฐ˜ ์ง€์—ญํ™”๊ฐ€ ๋ถ€์กฑํ•˜๊ณ , ๊ทธ๋ž˜ํ”„๊ฐ€ ๋ฌด๋ฐฉํ–ฅ์— ๋จธ๋ฌผ๋Ÿฌ ๋ฐฉํ–ฅ์„ฑ ํ† ํด๋กœ์ง€ ๋ฅผ ํ™œ์šฉํ•˜์ง€ ๋ชปํ•œ๋‹ค. | | ํ•ต์‹ฌ ๋ฌธ์ œ | ์ง€์—ญ๋ณ„(ETDRS, ODโ€‘fovea aligned) ๋ฐ”์ด์˜ค๋งˆ์ปค์™€ ํ† ํด๋กœ์ง€ ๊ธฐ๋ฐ˜ ์ง€ํ‘œ(๋ถ„๊ธฐ๊ฐ, ๋น„ํ‹€๋ฆผ ๋“ฑ)์˜ ์ •ํ™•ํ•˜๊ณ  ์žฌํ˜„ ๊ฐ€๋Šฅํ•œ ์ถ”์ถœ์ด ์–ด๋ ค์›€. | 2. Va

Quantitative Biology

< Category Statistics (Total: 2829) >

Electrical Engineering and Systems Science
100
General
731
General Relativity
22
HEP-EX
17
HEP-LAT
3
HEP-PH
39
HEP-TH
19
MATH-PH
36
NUCL-EX
2
NUCL-TH
5
Quantum Physics
41

Start searching

Enter keywords to search articles

โ†‘โ†“
โ†ต
ESC
โŒ˜K Shortcut