Intro to Computer Vision

NYU, Spring 2022

Source: Willow

Course webpage for the NYU Spring 2022 Course Special Topics in Data Science, DS-GA 3001.003/.004 (Introduction to Computer Vision). This course is aims to cover a broad topics in computer vision, and is not primarily a deep learning course. We will covert topics in traditional computer vision such as camera geometry, image formation, segmentation, object recognition, classification, and detection (see Syllabus).


  • DS-GA 3001.003 (Lecture)
    Thursdays 4:55pm-6:35pm
    Location: Global Center for Academic & Spiritual Life, 238 Thompson Street (GCASL 361)

    The class will be in-person or via Zoom (see Brightspace for link). Slides will be available after the class on this webpage. (see Schedule).

  • DS-GA 3001.004 (Lab)
    Tuesdays 5:55pm-6:45pm
    Location: Global Center for Academic & Spiritual Life, 238 Thompson Street (GCASL 361)

    The labs will be in-person or via Zoom (see Brightspace for link). Labs will be used to cover additional materials or to work through practical exercises with the TA.

  • Office Hours:
    Office hours with each instructor will be available by appointment.

  • Campuswire Link: Link
    Please try first to post any questions about course logistics and material, HWs and final project on Campuswire. We also highly encourage you to help each other out (but please do not reveal answers). For additional questions, please email course staff.


Jean Ponce (
Alberto Bietti (
Elena Sizikova (


Irmak Guzey (


Four programming assignments (50% of the grade) + final project (40% of the grade) + class participation and attendance (5%) + lab participation and attendance (5%). Assignments should be submitted using the Brightspace site.

  • Excercise 1 on camera calibration. Due on TBA.
  • Excercise 2 on Canny edge detector. Due on TBA.
  • Excercise 3 on mean shift. Due on TBA.
  • Excercise 4 on neural networks. Due on TBA.
  • Final project: details TBA.

Participation and Attendance

You are expected to attend and participate in classes and labs in person or via Zoom (see NYU Brightspace for link). Class attendance will count for 5% of your grade and lab attendance will count for 5% of your grade.


  • Introduction
  • Camera geometry and calibration
  • Filtering and feature detection
  • Radiometry and color
  • Texture and image segmentation
  • Stereopsis
  • Structure from motion and 3D models from images
  • Object recognition - historical perspective
  • CNNs for object classification and detection
  • 3D CNNs, Applications in Medical Imaging
  • Weakly-supervised and unsupervised approaches to image and video interpretation


We do not require purchase of any textbooks and the course will be self-contained. You may wish to consult the resources below for additional material formalization.

  • D.A. Forsyth and J. Ponce, “Computer Vision: A Modern Approach”, second edition, Pearson, 2011. (Link)
  • R. Szeliski, “Computer Vision: Algorithms and Applications”. (PDF)
  • R. Hartley and A. Zisserman, “Multiple View Geometry in Computer Vision”, Cambridge University Press, 2004. (Link)


Note: lecture slides will be posted after each lecture.

Date Lecture Topic Link
01/25 Lab: Course Intro Introduction, Logistics Slides (Keynote, PDF)
01/27 Lecture 1    
02/01 Lab    
02/03 Lecture 2    
02/08 Lab    
02/10 Lecture 3    
02/15 Lab    
02/16 Exercise 1 DUE    
02/17 Lecture 4    
03/22 Lab    
02/24 Lecture 5    
03/01 Lab    
03/03 Lecture 6    
03/08 Lab    
03/09 Exercise 2 DUE    
03/10 Lecture 7    
03/17 No class (spring break)  
03/22 Lab    
03/24 Lecture 8    
03/29 Lab    
03/31 Lecture 9    
04/05 Lab    
04/06 Exercise 3 DUE    
04/07 Lecture 10    
04/12 Lab    
04/14 Lecture 11    
04/19 Lab    
04/21 Lecture 12    
04/26 Lab    
04/28 Lecture 13    
05/03 Lab    
05/04 Exercise 4 DUE    
05/05 Lecture 14    


Much of the material for this course relies on the Computer Vision course given at ENS Paris by Mathieu Aubry, Karteek Alahari, Ivan Laptev, and Josef Sivic. Many of the slides are taken from James Hays, Svetlana Lazebnik, and Derek Hoeim. Website was originally designed by Matthew Trager.