1. Review of computer vision and deep learning concepts
- Definition of computer vision and related problems
- Image formation
- Brief overview of traditional methods
- Brief overview of deep learning methods
2. Object recognition
- Problem definition
- Traditional methods
- Convolutional neural networks for image classification (AlexNet, ResNet, DenseNet)
- Transformer networks for image classification (Vision Transformers)
- Evaluation of image classifiers
- Fine-tuning
- Python implementation of object recognition pipelines
3. Object detection
- Problem definition
- Traditional methods
- Two-stage detectors (R-CNN family)
- Single-stage detectors (YOLO)
- Efficient deep learning-based detectors
- Evaluation of object detectors
4. Object tracking
- Problem definition
- Traditional methods
- Offline vs Offline-online trackers
- Siamese network-based trackers
- Deep discriminative trackers
5. Final practice (homework) |