๋ณธ๋ฌธ ๋ฐ”๋กœ๊ฐ€๊ธฐ

๋ถ„๋ฅ˜ ์ „์ฒด๋ณด๊ธฐ31

[Git] local to github & directory๋ช… ๋ณ€๊ฒฝ ๋˜๋˜ ๊นŒ๋จน์„๊นŒ๋ด ์˜ฌ๋ฆฌ๋Š” local to github ๋ช…๋ น์–ด ์ •๋ฆฌ! $ git init // ๋กœ์ปฌ ์ €์žฅ์†Œ ์ƒ์„ฑ$ git status // ํ˜„์žฌ ํด๋” ํŒŒ์ผ๋“ค ํ™•์ธ// ์ „์ฒด ์—…๋กœ๋“œ or ํŠน์ • ํŒŒ์ผ/ํด๋” ์—…๋กœ๋“œ ์„ ํƒ$ git add . // ๋กœ์ปฌ ์ €์žฅ์†Œ์— ์ „์ฒด ํŒŒ์ผ ์—…๋กœ๋“œ$ git add [ํŒŒ์ผ๋ช…/ํด๋”๋ช…] // ๋กœ์ปฌ ์ €์žฅ์†Œ์— ํŠน์ • ํŒŒ์ผ/ํด๋” ์—…๋กœ๋“œ$ git commit -m "[ํ‘ธ์‰ฌ ๋ฉ”์‹œ์ง€]"$ git remote add origin [๋ณต์‚ฌํ•ด๋‘” ๋ ˆํฌ์ง€ํ† ๋ฆฌ ์ฃผ์†Œ] // github repo์—์„œ code > html ๋ณต๋ถ™$ git remote -v // ๋กœ์ปฌ ์ €์žฅ์†Œ๋ฅผ ์›๊ฒฉ ์ €์žฅ์†Œ๋กœ ์—ฐ๊ฒฐ$ git push origin master // ๋ฐฉ๊ธˆ ์ปค๋ฐ‹ํ•œ ํŒŒ์ผ๋“ค ์›๊ฒฉ์ €์žฅ์†Œ์— ์—…๋กœ๋“œ  ๋‚ด๊ฐ€ ์žŠ์„๊นŒ๋ด ์ ์–ด๋ณด๋Š” Git ํด๋”๋ช… .. 2024. 1. 11.
[Medical][Data] Creation and validation of a chest X-ray dataset with eye-tracking and report dictation for AI development, Sci Data 8, 92 (2021)(๋…ผ๋ฌธ๋ฆฌ๋ทฐ) โœจ ABSTRACTโ—‹  Developing a rich dataset : CXR image, transcribed radiology report text, radiologist's dictation audio and eye gaze coordinates data.โ—‹  Validating dataset by using deep learning experiments .  โœจ BACKGROUNDโ—‹  In the past five decades eye-tracking has been extensively used in radiology for education, perception understadning, and fatigue measurement.๋”๋ณด๊ธฐ1. Waite, S. A. et al. Analysis.. 2024. 1. 11.
[HCI] About Eye-Tracking... ๐Ÿ‘€ Eye movements๋Š” ํฌ๊ฒŒ ๋‘ ๊ฐ€์ง€(fixation, saccade) ํ˜น์€ ์„ธ ๊ฐ€์ง€(+smooth peusuit)๋กœ ๊ตฌ๋ถ„๋œ๋‹ค. Fixations are movements that occur when the eye momentarily stops to record visual data. Although the duration of different fixations varies, the longer the fixation, the more visual data is collected. And it can indicate quite a bit of meaningful information on the user’s attention and understanding. A shorter period of.. 2024. 1. 10.
[AI][Medical] Follow My Eye: Using Gaze to Supervise Computer-Aided Diagnosis (IEEE Trans Med Imaging. 2022 Jul) (๋…ผ๋ฌธ๋ฆฌ๋ทฐ) ์›๋ฌธ์ถœ์ฒ˜: https://ieeexplore.ieee.org/document/9694633 Follow My Eye: Using Gaze to Supervise Computer-Aided DiagnosisWhen deep neural network (DNN) was first introduced to the medical image analysis community, researchers were impressed by its performance. However, it is evident now that a large number of manually labeled data is often a must to train a properly functionieeexplore.ieee.org โœจ ABSTRA.. 2024. 1. 8.
[AI] What is ResNet? ๐Ÿ“ What is ImageNet? ImageNet challenge๋Š” Computer vision์—์„œ ์œ ๋ช…ํ•œ challenge๋‹ค. ImageNet์€ ์ˆ˜๋งŽ์€ ์ด๋ฏธ์ง€๋“ค์˜ ์ง‘ํ•ฉ์„ ๋ฐ์ดํ„ฐ๋ฒ ์ด์Šค์— ์ €์žฅํ•ด๋‘” ๊ฒƒ์œผ๋กœ, ์ด๋ฅผ classificationํ•˜๋Š” task๋กœ ๋ชจ๋ธ์˜ ์„ฑ๋Šฅ์„ ๊ฒจ๋ค„์™”๋‹ค. (*Top-5 error๋ž€? 1๊ฐœ๋‹น prediction ๊ฒฐ๊ณผ๋ฅผ 5๊ฐœ ์œ ์ถ”ํ•  ์ˆ˜ ์žˆ๊ฒŒ ํ•˜๊ณ  ๊ทธ 5๊ฐœ ์ค‘ ์ •๋‹ต์ด ์žˆ์œผ๋ฉด accurate ํ•œ ๊ฒƒ์œผ๋กœ ๊ฐ„์ฃผ.) ์œ„์˜ ๊ทธ๋ž˜ํ”„๋ฅผ ๋ณด๋ฉด 2012๋…„ AlexNet์œผ๋กœ ์œ ์˜๋ฏธํ•œ error ๊ฐ์†Œ๊ฐ€ ๋‚˜ํƒ€๋‚˜๊ณ , 2015๋…„ ResNet์œผ๋กœ ์ฒ˜์Œ ์ธ๊ฐ„์„ ๋›ฐ์–ด๋„˜๋Š” ์„ฑ๋Šฅ์„ ๋ณด์ธ ๊ฒƒ์„ ์•Œ ์ˆ˜ ์žˆ๋‹ค. ๐Ÿ“ What is ResNet? ๊ทธ๋ ‡๋‹ค๋ฉด ResNet์€ ๋ฌด์—‡์ผ๊นŒ? He, Kaiming; Zhang, Xiangyu.. 2024. 1. 5.
[Medical] Eye Motion Tracking for Medical Image Interpretation Training (Current Problems in Diagnostic Radiology, 2023) (๋…ผ๋ฌธ๋ฆฌ๋ทฐ) โœจ MAIN IDEAUSAF(United States Air Force)์—์„œ๋Š” ์‹ค์ œ๋กœ ๊ต์œก์„ ์œ„ํ•ด Eye Motion Tracking์„ ์‚ฌ์šฉํ•œ๋‹ค. ์ด๋ฅผ ์˜ํ•™ ๊ต์œก์— ๋„์ž…ํ•˜๊ณ ์ž, Novice reader(jounior residents)๋ฅผ ๊ต์œกํ•˜๊ธฐ ์œ„ํ•ด ์‹ค์‹œ๊ฐ„ Eye Motion Tracking์„ ์‚ฌ์šฉํ•˜๊ณ  ๊ทธ ํšจ๊ณผ๋ฅผ ๊ฒ€์ฆํ•˜๋Š” ์—ฐ๊ตฌ๋ฅผ ์ง„ํ–‰ํ–ˆ๋‹ค. ๋‹ค์‹œ๋งํ•ด ์ด paper๋Š” novice reader์˜ Eye Motion Tracking์„ expert๊ฐ€ ์‹ค์‹œ๊ฐ„ ํ”ผ๋“œ๋ฐฑํ•˜๋Š” ๊ต์œก ์‹œ์Šคํ…œ์„ ๊ตฌ์ถ•ํ•˜๊ณ  ๊ทธ ํšจ๊ณผ๋ฅผ ํ‰๊ฐ€ํ•œ ์—ฐ๊ตฌ๋ฅผ ์ •๋ฆฌํ•œ ๊ฒƒ์ด๋‹ค.  โœจ METHODS์‹ค์‹œ๊ฐ„ Eye Motion Tracking ์ „๋ฌธ๊ฐ€์˜ ๊ตฌ๋‘ ์ง€์‹œ๋ฅผ ํ†ตํ•ฉํ•˜์—ฌ ์ดˆ๋ณด์ž๋ฅผ ๊ต์œกํ•œ๋‹ค. ์‹คํ—˜์—๋Š” Dell Precision 7535 ์ปดํ“จํ„ฐ์™€ Gazepoint .. 2024. 1. 2.
[Python][AI] pandas, numpy ๊ธฐ์ดˆ ํ•จ์ˆ˜ ์ •๋ฆฌ array([[-1.5, 0.5], [ 1. , 0. ]])๊ธฐ์ดˆ์ธ๊ณต์ง€๋Šฅ ์ˆ˜์—… Project๋ฅผ ํ•˜๋ฉด์„œ pandas, numpy ํ•จ์ˆ˜๋ฅผ ์—„์ฒญ ์“ฐ๋Š”๋ฐ,์ •๋ฆฌํ•ด๋‘๋ฉด ์ข‹์„ ๊ฒƒ ๊ฐ™์•„์„œ ๋„์ ๋„์  ํ•ด๋ณด๋„๋ก ํ•˜๊ฒ ๋‹ค.  [Pandas] .dropna() Missing value๊ฐ€ ์žˆ๋Š” ํ–‰์„ ์‚ญ์ œํ•˜๋Š” ํ•จ์ˆ˜. A B C0 1.0 NaN 11 2.0 2.0 22 3.0 3.0 33 NaN 4.0 44 5.0 5.0 5๋งŒ์ผ ๋ฐ์ดํ„ฐ๊ฐ€ ์œ„์™€ ๊ฐ™์„ ๋•Œ data = data.dropna()๋ฅผ ํ•ด์ค€๋‹ค๋ฉด ์•„๋ž˜์™€ ๊ฐ™์ด ์ •๋ฆฌ๋œ๋‹ค. A B C1 2.0 2.0 22 3.0 3.0 34 5.0 5.0 5  [NumPy] concatenate()๋ฐฐ์—ด์„ ๊ฒฐํ•ฉ, ์—ฐ๊ฒฐํ•ด์ฃผ๋Š”.. 2023. 10. 9.
[HCI] Bounded Rationality ์ œํ•œ๋œ ํ•ฉ๋ฆฌ์„ฑ, ํœด๋ฆฌ์Šคํ‹ฑ ์ธ๊ฐ„์€ ๊ณผ์—ฐ ํ•ฉ๋ฆฌ์ ์ผ๊นŒ? ์ธ๊ฐ„๊ณผ ์ปดํ“จํ„ฐ ์ธํ„ฐํŽ˜์ด์Šค ์ˆ˜์—…์—์„œ ๊ต์ˆ˜๋‹˜์ด ๋˜์ง€์‹  ์งˆ๋ฌธ์ด๋‹ค. ์ธ๊ฐ„์ฒ˜๋Ÿผ ํ–‰๋™ํ•˜๊ณ  ์ธ๊ฐ„์ฒ˜๋Ÿผ ๋Œ€๋‹ตํ•˜๋Š” ๊ธฐ๊ณ„๋ฅผ ๋งŒ๋“œ๋ ค๋Š” ์ˆ˜๋งŽ์€ ๋…ธ๋ ฅ์ด ์žˆ์–ด์™”๊ณ , ํ˜„์žฌ๋„ ์ง„ํ–‰ ์ค‘์ด๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๊ทธ๋“ค์ด ์ซ“๋Š” ์ธ๊ฐ„์€ ๊ณผ์—ฐ ํ•ฉ๋ฆฌ์ ์ผ๊นŒ? ์ด์— ๋Œ€๋‹ต์ด ๋˜๋Š” ์ด๋ก ์ด ์žˆ๋Š”๋ฐ, ๋ฐ”๋กœ '์ œํ•œ๋œ ํ•ฉ๋ฆฌ์„ฑ' ์ด๋ก ์ด๋‹ค. ์ œํ•œ๋œ ํ•ฉ๋ฆฌ์„ฑ์€ ๊ฐœ์ธ์ด ๊ฒฐ์ •์„ ๋‚ด๋ฆด ๋•Œ ํ•ฉ๋ฆฌ์„ฑ ์ด ์ œํ•œ๋œ๋‹ค๋Š” ์ƒ๊ฐ์ด๋‹ค. ์ฆ‰, ์ธ๊ฐ„์˜ ์„ ํ˜ธ๋„๋Š” ํŠน์ • ๊ธฐ์ค€ ์ˆ˜์ค€์— ๋Œ€ํ•œ ๊ฒฐ๊ณผ์˜ ๋ณ€ํ™”์— ์˜ํ•ด ๊ฒฐ์ •๋œ๋‹ค๋Š” ๊ฒƒ์ด๋‹ค. ์ด ์ด๋ก ์€ Herbert A. Simon์— ์˜ํ•ด ๋งŒ๋“ค์–ด์กŒ์œผ๋ฉฐ, ๋‹ค์–‘ํ•œ ๋ถ„์•ผ์—์„œ ์‚ฌ์šฉ๋˜๋Š” ์˜์‚ฌ ๊ฒฐ์ •์˜ ์ˆ˜ํ•™์  ๋ฐ ์‹ ๊ณ ์ „์ฃผ์˜์  ๊ฒฝ์ œ ๋ชจ๋ธ๋ง์„ ์œ„ํ•œ ๋Œ€์•ˆ์  ๊ธฐ๋ฐ˜์œผ๋กœ ์ œ์•ˆ๋˜์—ˆ๋‹ค. Simon์€ ๊ฒฝ์ œ ์ฃผ์ฒด๊ฐ€ ์—„๊ฒฉํ•˜๊ณ  ์—„๊ฒฉํ•œ ์ตœ์ ํ™” ๊ทœ์น™๋ณด๋‹ค ๊ฒฐ์ •์„ ๋‚ด๋ฆฌ๊ธฐ ์œ„ํ•ด ํœด๋ฆฌ์Šคํ‹ฑ ์„.. 2023. 9. 8.