h1
Jaejun Yoo's Playground
h2
#338. Alias-Free Generative Adversarial Networks
h2
#323.Separation and Concentration in Deep Networks
h2
#385. Generative Modeling by Estimating Gradients of the Data Distribution
h2
#374. Fourier Features Let Networks Learn High-Frequency Functions in Low Dimensional Domains
h2
#312.Generative Models as Distributions of Functions
h2
Properties of Orthonormal Bases
h2
To be continued ... (planned)
h2
Signals and Hilbert Spaces
h2
From Vector Spaces to Hilbert Spaces
h2
Inner Products and Distances
h2
To be continued ... (planned)
h2
Discrete-time 신호의 종류 네 가지
h2
How sensitive is my system?
h2
System with single input and output variables
h2
System with multiple input and output variables
h2
Another representation
h2
#87.Spectral Normalization for Generative Adversarial Networks
h2
To be continued ... (planned)
h2
Single Image Super-Resolution
h2
SRCNN: The Start of Deep Learning in SISR
h3
How sensitive is my system?: Condition number (조건수)
h3
Deep Learning for Super-Resolution: A Survey (1)
h3
[PR12-Video] 87. Spectral Normalization for Generative Adversarial Networks
h3
Infinite-length signals: Aperiodic
h3
Infinite-length signals: Periodic
h3
Infinite-length signals: Finite-support
h3
The Reproducing Formula
h3
Finite Euclidean Spaces
h3
Signal Processing For Communications (2)
h3
Finite Euclidean Spaces.
h3
Square Summable Functions.
h3
Signal Processing For Communications (3-1)
h3
Synthesis and Analysis Formula.
h3
Best Approximations (Projections).
h3
Orthogonal/Orthonormal Basis.
h3
Signal Processing For Communications (3-2)
h3
복소 함수의 다가성 (multivaluedness of complex function)
h3
[PR12 Video] 338. Alias-Free Generative Adversarial Networks (StyleGAN3) 리뷰 개념 정리
h3
[PR12 Video] 374. Fourier Features Let Networks Learn High-Frequency Functions in Low Dimensional Do
h3
[PR12 Video] 312. Generative Models as Distributions of Functions
h3
[PR12 Video] 323. Separation and Concentration in Deep Networks
h3
[PR12 Video] 385. Generative Modeling by Estimating Gradients of the Data Distribution
h4
수열 $x[n]$를 factor $a\in\mathbb{C}$만큼 scaling하는 것은 $$y[n]=ax[n]$$이라 표현됩니다. 여기서 a가 실수이면 신호의 amplificat
h4
수열 $x[n]$와 수열 $w[n]$을 더하거나 곱하는 것은 elementwise로 수행합니다: $$y[n]=x[n]+w[n], \quad y[n]=x[n]w[n]$$
h4
Discrete-time에서의 integration은 다음과 같이 합으로 정의됩니다: $$y[n]=\sum_{k=-\infty}^{n}x[k]$$