Computer Vision

176 terms

Foundations & Concepts

15 terms

Aspect Ratio
Bit Depth
Bounding Box Annotation Format
Color Channels
Frame Rate
Ground Truth Mask
Image Annotation
Image as Tensor
Image Format
Image Resolution
Pixel
Region of Interest (ROI)
Sliding Window
Spatial Coordinates
Video Frame

Image Preprocessing

18 terms

Spatial / Geometric Transforms 6

Center Crop
Horizontal Flip
Image Resize
Random Crop
Random Rotation
Vertical Flip

Pixel / Colour Transforms 7

Channel Order (BGR vs RGB)
Color Space (RGB, HSV)
Grayscale Conversion
Histogram Equalization
Image Normalization
Mean Subtraction
Pixel Scaling

Augmentation Strategies 5

AutoAugment
Color Jitter
Cutout
Gaussian Blur (Augmentation)
RandAugment

Classical CV

22 terms

Edge & Gradient Operators 5

Canny Edge Detection
Laplacian of Gaussian (LoG)
Prewitt Operator
Sobel Operator
Image Gradient

Feature Detectors & Descriptors 9

Corner Detection (Harris)
Histogram of Oriented Gradients (HOG)
Oriented FAST and Rotated BRIEF (ORB)
Scale-Invariant Feature Transform (SIFT)
Speeded-Up Robust Features (SURF)
BRIEF Descriptor
FAST Corner Detector
Viola-Jones Detector
Haar cascades

Matching & Tracking 4

Lucas-Kanade Method
Optical Flow
Template Matching
Keypoint Matching

Multi-Scale & Aggregation 4

Bag of Visual Words
Hough Transform
Image Pyramid
Scale-Space

CNN Architectures

19 terms

Foundational Architectures 4

AlexNet
LeNet
Network in Network
VGG Net

Inception Family 4

GoogLeNet (Inception)
Inception Module
Xception
Inception-v3

ResNet & Dense Family 4

DenseNet
Residual Block
ResNet
ResNeXt

Efficient Architectures 7

ConvNeXt
EfficientNet
MobileNet
NASNet
SqueezeNet
MobileNetV2
ShuffleNet

Recognition Tasks

12 terms

Face Detection
Face Recognition
Facial Landmark Detection
Fine-grained Recognition
Image Classification
Image Retrieval
Keypoint Detection
Multi-Label Image Classification
Optical Character Recognition (OCR)
Person Re-identification
Pose Estimation
Scene Classification

Object Detection

21 terms

Detection Concepts 8

Anchor Box
Average Precision (AP)
Bounding Box
Feature Pyramid Network (FPN)
Intersection over Union (IoU)
Mean Average Precision (mAP)
Non-Maximum Suppression (NMS)
Region Proposal

Two-Stage Detectors 4

Fast R-CNN
Faster R-CNN
R-CNN
Cascade R-CNN

Single-Stage & Modern Detectors 9

DETR (Detection Transformer)
RetinaNet
SSD (Single Shot Detector)
YOLO (You Only Look Once)
YOLOv3
Deformable DETR
YOLOv5
YOLOv7
YOLOv8

Segmentation

16 terms

Segmentation Tasks 4

Instance Segmentation
Panoptic Segmentation
Pixel-wise Classification
Semantic Segmentation

Segmentation Architectures 8

Atrous Spatial Pyramid Pooling (ASPP)
DeepLab
Fully Convolutional Network (FCN)
Mask R-CNN
Pyramid Pooling Module
SegNet
U-Net
Segment Anything Model (SAM)

Segmentation Losses & Metrics 4

Dice Loss
IoU Loss
Mean Intersection over Union (mIoU)
Pixel Accuracy

Transfer Learning

11 terms

Backbone
Backbone Fine-tuning
Detection Head
Discriminative Learning Rates
Domain Adaptation
Frozen Feature Extraction
ImageNet Pretraining
Knowledge Distillation
Layer Freezing
Pretrained Model
Progressive Unfreezing

Vision Transformers

19 terms

Core ViT 5

CLS Token
Image Patches
Patch Embedding
Vision Transformer (ViT)
Class Token Pooling

ViT Variants 5

BEiT
DeiT (Data-efficient Image Transformer)
Hybrid CNN-Transformer
Shifted Windows
Swin Transformer

Self-Supervised Visual Learning 9

DINO (Self-Supervised)
Masked Autoencoder (MAE)
BYOL
MoCo
SimCLR
CLIP
ALIGN
OpenCLIP
DINOv2

Video Understanding

11 terms

Action Recognition
C3D
Frame Sampling
I3D (Inflated 3D ConvNet)
Multi-Object Tracking (MOT)
SlowFast Networks
Temporal Convolution
Temporal Pooling
Two-Stream Network
Video Classification
Video Object Segmentation

3D Vision

12 terms

3D Mesh
3D Reconstruction
Depth Estimation
Disparity Map
Neural Radiance Fields (NeRF)
Point Cloud
PointNet
SLAM
Stereo Vision
Structure-from-Motion (SfM)
Visual Odometry
Voxel Grid