Quiz 4 Topic List (tentative)

Introduction to Recognition

Machine Learning Basics

1. Nearest neighbors

2. Support Vector Machines

3. Support Vector Machines with Kernels

4. Bias variance trade-off

5. Model Selection

Feature Engineering

1. Digit Classification Case Study

Deep Neural Networks

1. Back-propagation

2. Development Cycle

3. Hyper-parameters and their affect

Image Classification

1. Architectural Constructs and their Properties

   1. Convolutions

   2. Pooling

   3. Fully-connected Layers

   4. Non-linearities

   5. Dropout

   6. Batch-normalization

   7. Residual Connections

   8. Attention

2. Popular Architectures

3. Using ImageNet pre-trained models for vision tasks

Segmentation

1. Tasks and Metrics

   1. Bottom-up segmentation

   2. Semantic segmentation

   3. Instance segmentation

2. Shift and Stitch, Dilated Convolutions

3. Skip-connections

4. Encoder-decoder architectures

Detection

1. Problem definition

2. Metrics

3. Basics of classical approaches

4. CNN based Detectors

   1. R-CNN

   2. Fast R-CNN

   3. Faster R-CNN

   4. Single-shot Detectors

   5. Feature Pyramid Networks

5. ROI Pooling

6. Non-Maximum Suppression

7. Region Proposal Networks

8. Instance Segmentation

   1. Metrics

   2. RoI Align vs RoI Pooling

Transformers

1. Attention

2. Vision Transformers

3. Visual Prompt Tuning

4. Swin Transformers

5. Plain ViT backbones for detection

6. DETR

Self-supervision

1. MAE