ConceptMathematics
Convolution
An operation on two functions that produces a third function, representing the distribution of the sum of two independent random variables.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Opens up
Concepts that name this one as a prerequisite.
Circulant matrix2Convolutional neural networks2Cyclic convolution2Deconvolution2Activation map1Anti-aliasing filters1Bilateral filtering1Camera shake as convolution1Complex cell energy models1Convolution identity property of delta function1Convolution of log-concave functions1Convolution of probability density functions1
Lecturers
13Each in their own terms, from the easiest level.
Joe Blitzstein
The distribution of the sum of two independent random variables.Gilbert Strang
An operation on two functions or vectors resulting in a third, representing how the shape of one is modified by the other.Josh McDermott
An operation that produces the output of a linear filter at each location in an image or every time point in some time series.Sara Beery
A linear shift-invariant transformation of grid-structured input data used as filtering in signal processing.Arthur Mattuck
An operation on two functions that produces a third function, corresponding to the product of their Laplace transforms.Erik Demaine & Srini Devadas
An operation on two vectors that computes the dot product of one with all possible shifts of the other.Andrej Karpathy
Operations that slide over an input sequence to calculate all output positions simultaneously.Patrick Winston
Running a neuron (kernel) across subsets of an image to produce an output map.Fei-Fei Li, Justin Johnson & Serena Yeung
A mathematical operation between a signal and a flipped version of a filter, implemented in convolutional neural networks by sliding a filter across spatial locations.Onur Mutlu
A fundamental filtering operation used in image processing and neural networks to compute a weighted sum of input sequences.Marc Levoy
Replacing each input value with a weighted sum of itself and its neighbors using a filter function.Stephen Boyd
An operation on matrix-valued functions represented by the star symbol, used to describe the output of a linear dynamical system as an integral over past inputs weighted by the impulse response.
Lectures that assume it
22The first 12 are shown.
- Convex Optimization Problems — LP, QP, SOCP & SDP
- Harvard CS50’s Artificial Intelligence with Python – Full University Course
- Applications of Least-Squares Estimation: System Identification & Filtering
- CNN Architectures
- Detection and Segmentation
- Convex Optimization with Abstract Linear Operators
- Texture Perception (cont'd)
- Gibbs Measures on Infinite Trees and Homoclinic Points
- Håstad's Hardness Theorems
- Impulse Inputs: Dirac Delta Function, Weight and Transfer Functions
- Analysis of Boolean Functions: Additive Combinatorics
- Sanders's Theorem (Analysis of Boolean Functions, CMU 15-859S, Lecture 22)
Connections
51The lectures that define it or assume it, and the concepts next to it.
Concepts10
Show 4 moreShow fewer
Lectures41
- Divide & Conquer: FFT
- Building makemore Part 5: Hierarchical Dilated Causal Convolutions (WaveNet)
- Architectures: Memory
- Lateral Geniculate Nucleus (LGN) / Primary Visual Cortex (V1)
- Moment Generating Functions
- Convolution Formula: Proof, Connection with Laplace Transform, Application to Physical Problems
Show 35 moreShow fewer
- Probability Densities and BLR Linearity Testing
- Completing a Rank-One Matrix and Circulants
- Transformations and Convolutions
- The Eye and the Retina (cont'd)
- Deep Neural Networks: MIT Artificial Intelligence
- Autofocus Systems: Active, Contrast, and Phase Detection (CS 178, Lecture 7)
- Convolutional Neural Networks
- VLIW and Systolic Array Architectures
- Eigenvectors of Circulant Matrices: Fourier Matrix
- ImageNet and Convolutional Neural Networks
- Architectures: Grids
- Applications of Jordan Canonical Form in LDS and Electrical Engineering
- MGFs Continued
- Convex Optimization Problems — LP, QP, SOCP & SDP
- Harvard CS50’s Artificial Intelligence with Python – Full University Course
- Applications of Least-Squares Estimation: System Identification & Filtering
- CNN Architectures
- Detection and Segmentation
- Convex Optimization with Abstract Linear Operators
- Texture Perception (cont'd)
- Gibbs Measures on Infinite Trees and Homoclinic Points
- Håstad's Hardness Theorems
- Impulse Inputs: Dirac Delta Function, Weight and Transfer Functions
- Analysis of Boolean Functions: Additive Combinatorics
- Sanders's Theorem (Analysis of Boolean Functions, CMU 15-859S, Lecture 22)
- Progress on Zeta and L-Functions Motivated by the Riemann Hypothesis
- Learning Representations: A Challenge for Learning Theory, COLT 2013
- Course Overview: Applied Linear Algebra & Linear Dynamical Systems
- Deep Learning Software
- Object Recognition (cont'd), Texture Perception
- Autofocus Systems and Camera Mechanics
- Architectures: Transformers
- Lectures on Digital Photography: Lecture 4 (Stanford CS 178)
- Color Sensing and Human Vision
- Digital Photography: Light Fields and Computational Cameras