CMPSCI 590IV + 690IV:
Intelligent Visual Computing

teaser


Time: TBA
Location: TBA
Instuctor: Evangelos Kalogerakis (kalo@cs.umass.edu)
Office hours: Mondays 3-5pm, office CS250; at other times, drop in or make an appointment

Lecture notes, programming resources, and assignments will be posted on the UMass Moodle web site.

Course Description

Intelligent visual computing is an emerging new field that seeks to combine modern trends in machine learning, computer graphics, computer vision and human-computer interaction to intelligently process, analyze and synthesize 2D/3D visual data.

The course will start by covering image and shape representations, basic classification and regression techniques, and the fundamentals of deep learning. The course will then provide an in-depth background on analysis and synthesis of images and shapes with machine learning. Particular emphasis will be given on discriminative and generative probabilistic models of visual data, convolutional neural networks, recurrent neural networks, auto-encoders, and adversarial networks.

Students will complete 5 programming assignments in Matlab/Octave and work on a course project related to visual computing with machine learning.

There are no prerequisites for CMPSCI graduate students, although familiarity with probability, statistics, linear algebra, undergraduate computer graphics, computer vision and machine learning is essential!

For others and undergrads, the course has the following prerequisites:
- Introduction to Algorithms (311 or equivalent) with grade B or better
- Artificial intelligence (383 or equivalent) with grade B or better
- Introduction to Computer Graphics (373, 473 or equivalent) with grade B or better

This course counts as a CS Elective toward the CMPSCI major (BA/BS) and can satisfy a second or third AI core requirement in the CMPSCI PhD and MSc program. It offers 3 credits.

Week-by-week outline (tentative)

Week 1: Introduction, Image and Shape Representations, Basics of Classification & Regression [Assignment 1]
Week 2: Image and Shape Descriptors
Week 3: 3D Object Reconstruction techniques [Assignment 2]
Week 4: Neural Networks, Convolutional Networks, Applications to Image Recognition
Week 5: Residual Nets, Siamese Nets, Autoencoders, Fully Convolutional Nets, Applications to Image Segmentation [Assignment 3]
Week 6: Recurrent Neural Networks, LSTMs Applications to Image and Video Analysis
Week 7: 3D Deep Learning [Assignment 4]
Week 8: Probabilistic Graphical Models, Applications to Image and Shape Analysis
Week 9: Inference in Probabilistic Graphical Models, Applications to 3D Modeling
Week 10: Learning Probabilistic Graphical Models, Image and Shape Priors [Assignment 5+6]
Week 11: Latent variables, Combining Probabilistic Graphical Models with Convnets
Week 12: Research paper presentations
Week 13: Project presentations

Requirements

590IV:
-
Complete six assignments in Matlab.
- Research paper presentation.
- Class participation and discussions.

690IV:
-
Complete five assignments in Matlab.
- A term project. This should be an implementation of a machine learning algorithm related to visual computing. The result of the project should be a prototype with some preliminary results, or could lead to a submission of a research paper in the future.
- Class participation and discussions.

Marking scheme

590IV:
10% Assignment 1 (warm-up): Shape Classification
15% Assignment 2: Deep neural networks for Object Recognition in Images
10% Assignment 3: 3D Shape Retrieval with Multi-View Convolutional Networks
15% Assignment 4: Surface Reconstruction from Point Clouds
15% Assignment 5: Feature-Preserving Surface Denoising with Markov Random Fields
10% Assignment 6: Laplacian Mesh Processing
15% Research paper presentation
10% Class participation

690IV:
10% Assignment 1 (warm-up): Shape Classification
15% Assignment 2: Deep neural networks for Object Recognition in Images
10% Assignment 3: 3D Shape Retrieval with Multi-View Convolutional Networks
15% Assignment 4: Surface Reconstruction from Point Clouds
15% Assignment 5: Feature-Preserving Surface Denoising with Markov Random Fields
25% Final project
10% Class participation

Scores to letter conversions:
95%-100% A
90%-95% A-
85%-90% B+
80%-85% B
75%-80% B-
70%-75% C+
65%-70% C
60%-65% C-
55%-60% D+
50%-55% D

Accommodation Statement

The University of Massachusetts Amherst is committed to providing an equal educational opportunity for all students.  If you have a documented physical, psychological, or learning disability on file with Disability Services (DS), you may be eligible for reasonable academic accommodations to help you succeed in this course.  If you have a documented disability that requires an accommodation, please notify me within the first two weeks of the semester so that we may make appropriate arrangements.

Academic Honesty Statement

Since the integrity of the academic enterprise of any institution of higher education requires honesty in scholarship and research, academic honesty is required of all students at the University of Massachusetts Amherst.  Academic dishonesty is prohibited in all programs of the University.  Academic dishonesty includes but is not limited to: cheating, fabrication, plagiarism, and facilitating dishonesty.  Appropriate sanctions may be imposed on any student who has committed an act of academic dishonesty.  Instructors should take reasonable steps to address academic misconduct.  Any person who has reason to believe that a student has committed academic dishonesty should bring such information to the attention of the appropriate course instructor as soon as possible.  Instances of academic dishonesty not related to a specific course should be brought to the attention of the appropriate department Head or Chair.  Since students are expected to be familiar with this policy and the commonly accepted standards of academic integrity, ignorance of such standards is not normally sufficient evidence of lack of intent (http://www.umass.edu/dean_students/codeofconduct/acadhonesty/).

 


back to Evangelos Kalogerakis' page