Academic Year 2020 - 2021
Teaching Assistant: Nachwa Aboubakr
Verify Lecture Locations in ADE
Schedule of Classes
http://crowley-coutaz.fr/jlc/Courses/2020/PRML/ENSI3.PRML.html
Class notes (pdf)
7 October 2020 Performance Evaluation for
Machine Learning
Recording of Lecture
Information about OpenCV and Jupyter Notebooks in
Conda
Python for lab exercises
UMASS FDDB: Face Detection Data Set and
Benchmark
Background
Reading: [Jain and Learned-Miller 2010]
14 October 2020 Detecting Faces in
Images
Recording of Lecture
Lab
Project 1: Evaluating the Viola Jones
Face Detector
(Oral
Presentation 4 November, Written reports due Wednesday 11
November)
21 October 2020 The Viola Jones
Face Detector
Recording of Lecture for Lesson
3
Background
Reading: Viola Jones 2001 CVPR Paper,
The
Viola Jones Faces Detector
28 October - No Class - Vacance de Toussaint
4 November 2020 Oral Presentations for project 1
11 novembre 2020 No
Class - Journée éthique et developpeent durable
Project 1 written reports due by
email.
Example of good report
for evaluation of the viola-jones Face detector
18 novembre 2020 No Class - programming project
25 November 2020 Perceptron
Learning and Gradient Descent
Recording
of Lecture
Lab Project 2: Face Detection with fully
connected neural networks
(Project due-date 23 December 2020)
02 December 2020 MultiLayer Artificial
Neural Networks, Backpropagation
Recording of Lecture
Information about Keras
and Tensorflow
Nachwa
Aboubakr's Tutorial on Keras
Paper on Xavier GLORIOT's initialisation
procedure
Background Reading: Face
Detection with Neural networks (Rowley-Kanade 1998)
09 December 2020 Encoders: Generative Networks, PCA, Auto-Encoders,
Variational Autoencoders and Generative Adversarial Networks.
Recording of Lecture
Example of Eigenspace Coding for Video Transmission (32 bytes per image)
Background Reading: Turk and Pentland, Face
recognition using eigenfaces, CVPR '91
Background Reading: Kingma
et al, Deep Generative Models, NIPS 2014
Background Reading: Review on Generative Adversarial Networks,
Gui et al 2021
16 December 2020 Oral Presentations for Lab 2.
23 December 2020 Written Report for Lab 2 due by email
Examples of Excellent Reports: Example1, Example2, Example3
06 January 2021 Convolutional Neural Networks,
Pooling, Classic Architectures
Recording of Lecture
Background Reading: VGG, Simonyan and
Zisserman, ICLR 2015
Installation for Conda Python with Keras and OpenCV
Simple CNN to test installation - classify MNist Digits - the MNIST Data Set
Lab 3: Face Detection with
CNNs - Written report due on 3 Feb 2021
13 January 2021 Generative Convolutional Networks, Deconvolution and YOLO: You Only Look
Recording of Lecture
Background Reading: Radford, A et al. Paper on DCGAN
Background Reading: Noh et
al, Deconvolution for Semantic Segmentation. ICCV 2015
Background Reading: J. Redmon, et al, You only look once: Unified, real-time object detection. CVPR 2016
Background Reading: J. Redmon et al, YOLO9000: Better, Faster, Stronger, CVPR 2017
20 January 2021 Support Vector Machines and Kernel Methods
SVM Tutorial by Shi Hung WU (CYUT)
Background Reading: Osuna, E. Freund, R. Girosi, F., Training Support Vector Machines: An Application to Face Detection. CVPR 97.
27 January 2021 Project Team
Presentations Lab 3
3 February 2021 Written reports for Lab project 3 due
by email
8 February 2021 Final Exam. (50% of grade)
Annals:
Exam from
Jan 2017
Exam from Feb 2018
Exam from Feb 2019
Exam from Feb 2020
Face Detection Data Sets:
FDDB dataset:
FDDB dataset contains the annotations for 5,171 faces in a set of 2,845
images.
WIDER FACE:
A face detection benchmark dataset with 32,203 images and labels for
393,703 faces with a high degree of variability in scale, pose and
occlusion.
Head
Pose Data Set: Created by Nicholas Gourier in 2004.
MALF dataset:
Face Detection in the Wild. MALF consists of 5,250 images and
11,931 faces.
LFW: Labeled
Faces in the Wild
IJB-A
dataset: US NIST IJB-A dataset for face detection and face
recognition. IJB-A contains 24,327 images and 49,759 faces.
http://crowley-coutaz.fr/jlc/Courses/2020/PRML/PRML.html
NB : Ces fichiers peuvent etre
copié, reproduit et
et inclus
dans autre texte, sous condition d'inclure une citation :
These files can be copied and used in editing other text, with the
condition of including a citation:
Copyright:
Crowley, J. L. , "Class Notes - Pattern Recognition and Machine
Learning", ENSIMAG, Grenoble Institut Polytechnique,
Crée par James L.
Crowley. Last update 1 Jan 2021