清華大學計科所
人工智慧
Artificial Intelligence
Robotic AI Lab
ICMS National Tsing Hua University
劉晉良
Jinn-Liang Liu

 

2020 Fall Course (Seminar)

Machine Learning for Self-Driving Cars

ENG1工一 209,  1:20 – 3:10 (3:20 – 5:20) pm, Wednesdays (2020.9.16 – 2021.1.6)

Send reports to jinnliu@mail.nd.nthu.edu.tw (Email Subject: Name, Student ID No., Report 1, 2, or 3.)

Report 1: ppt (demo) by 10/21. Report 2: ppt by 11/25. Report 3: ppt and docx (demo) by 1/6

Presentations on 12/31 and 1/6. *comma coding

 

 

Lecture Notes

l   *AI abc: An Introduction to Machine Learning

l   Gradient Descent and Backpropagation in Machine Learning (Automatic Differentiation: Forward & Reverse Modes, Jacobian)

l   Convolution in Machine Learning (Convolution)

l   Batch Normalization

 

Part I   Supervised Learning

1.        A Simple Learning Model: Classification, Target, Hypothesis, Training Data, Learning Algorithm, Weights, Bias, Supervised and Unsupervised Learning

2.        Google Tutorial for ML Beginners: Image Recognition, MNIST, Softmax Regression (92%), Cross Entropy, Gradient Descent, Back Propagation, Computation Graph (TF mnist 1.0)

3.        *Tensorflow and Deep Learning I (by Martin Gorner): Deep Learning (98%), ReLU, Learning Rate, Overfitting, Dropout (98.2%), Convolutional Neural Network (CNN, 99.3%)  (TF mnist 3.1)

4.        *Tensorflow and Deep Learning II (by Martin Gorner) (RNN1): Batch Normalization (99.5%) (TF mnist 4.2), MNIST Record (Kaggle: 100%), Recurrent Neural Network, Deep RNN, Long Short Term Memory, Gated Recurrent Network

 

Part II   Self-Driving Cars

1.        Introduction to Self-Driving Cars

l  Carnegie Mellon U 1989,  CMU Vehicle, Computer

l  comma.ai openpilot 2018, commaai, comma-GitHub, openpilot

l  Tesla Autopilot 2019, auto vs open

l  Self-Driving Car, Autonomous Car

2.        End-to-End Learning for Autonomous Driving

l  Video, Poster, *Paper

3.        Project 1: Steering Angle

l  *comma coding

l  Toyota Dynamic Radar Cruise Control, Adaptive Cruise Control,

4.        Project 2: Lane Detection

l  Code 2, Data 2, TuSimple

l  Toyota Lane Tracing Assist, Lane Centering

5.        Project 3: Speed Prediction

l  Code 3, Data 3, TuSimple

6.        Project 4: Localization

l  Code 4, Data 4, Laika

l  Theory: GNSS Processing, Trilateration, Least Squares,

7.        Driving Video Dataset

l  comma Data

l  Udacity Data

l  Nvidia Data

l  Berkeley Data, GitHub

8.        Hardware and Software in Self-Driving Cars

9.        Longitudinal and Lateral Control

10.    CAN Bus Protocol

11.    Environment Perception

12.    Traffic Signs Detection

13.    Pedestrian Detection

14.    How to ensure the safety of Self-Driving Cars

 

 

Publication

D.-H. Lee, K.-L. Chen, K.-H. Liou, C.-L. Liu, J.-L. Liu, Deep learning and control algorithms of direct perception for autonomous driving, Applied Intelligence (2020). (Video, Poster, Code and Data)

 

Past Courses