• University of Beira Interior


2021/22, Spring


- The Four "R"s of Computer Vision

- Geometry and Image Formation

- Light and Color, Cameras and Optics, Pixels and Image Representations

- Image Processing

- Linear Systems, Spatial and Frequency Domains, Convolution and Filters, Edge detection, Keypoints Detection

- Camera Calibration

- Epipolar Geometry

- Object Detection and Segmentation

- AdaBoost detector, Hough Transform, Active Contours

- Multiple Views and Motion

- Stereo Correspondence, Optical Flow

- Image Representation

- Feature Extraction, Principal Components Analysis

- Image Classification and Recognition

- Feature Matching, Nearest Neighbour Classification, Linear Classification, Support Vector Machines

- Neural Networks

- Cost Functions Optimisation, Gradient Descend, Retropropagation Algorithm

- Deep Learning Models

- Layers and Configurations, Convolutional Neural Networks, AlexNet, VGGNet, ResNet, Preprocessing, Data Augmentation


- R. Szeliski. Computer Vision: Algorithms and Applications. Springer, ISBN: 978-1848829343, 2021.

- E. R. Davies. Computer Vision: Principles, Algorithms, Applications, Learning. Academic Press, ISBN: 978-0128092842, 2018.

- D. Forsyth and J. Ponce. Computer Vision: A Modern Approach (2nd Edition), Pearson Publishing, ISBN: 978-0136085928, 2012.


- Assiduity (A) To get approved at this course, students should attend to - at least - 80% of the theoretical and practical classes

- Practical Project (P) The practical projects of this course weights 50% (10/20) of the final mark

- Dataset source files

Signatures M11324: [Shared Folder]

Signatures M11408: [Shared Folder]

Signatures M11467: [Shared Folder]

Signatures M11946: [Shared Folder]

Signatures M11819: [Shared Folder]

Signatures M11364: [Shared Folder]

Signatures M11911: [Shared Folder]

Signatures M11954: [Shared Folder]

Signatures M11466: [Shared Folder]

Signatures M12113: [Shared Folder]

- To get approved at the course, a minimal mark of 5/20 should be obtained in the practical project part;

- The pratical project mark is conditioned to an individual presentation and discussion by each student;

- Written Test (F) Wednesday, June 8th, 2022, 14:00. Room 6.18

- Mark (M) M = (A >= 0.8) * (P * 10/20 + F * 10/20)

- Admission to Exams Students with M >= 6 are admitted to final exams

- The practical projects mark is considered in all exam epochs;


Theoretical slides: [pdf]

Practical exercises (Introduction + Python essentials): [pdf]

Practical Work (Signature Recognition): [pdf]

Theoretical/Practical slides (Sample CV Problem): [pdf]

Theoretical/Practical slides: (Geometry Fundamentals) [pdf]

Theoretical slides (Calibration): [pdf]

Theoretical slides (Signals and Systems): [pdf]

Theoretical slides (Image Features): [pdf]

Practical sheet (OCR): [pdf]

Dataset OCR (imgs) [zip]

Theoretical slides (Deep Learning): [pdf]

Theoretical slides (Object detection): [pdf]

Practical Sheet (Object detection): [pdf]

GTSRB Dataset (Object detection): [zip]

(Consolidation lecture)

Theoretical slides (Semantic Segmentation / Loss Functions): [pdf]

Old Written Tests (Examples): [zip]

Classifier Script (Running Example): [py]

"CSV" Generator Script (Running Example): [py]

Theoretical slides (Deep Learning, cont): [pdf]

Theoretical slides (Deep Learning, Architectures): [pdf]

Theoretical slides (Gradient Descend + Backpropagation): [pdf]




Informatics Department

Theoretical + Practical classes