# Lecture 'Multimodal Sensor Systems'

## Course dates

All lecture and exercise dates.

## Slides

Basic termsIntroduction to CNN

Objektdetektion mittels CNNs (R-CNN, Fast R-CNN, Faster R-CNN)

Analyse von Machine Learning Modellen

Rekursives Bayes'sches Filtern

Partikelfilter

## Exercises

1. First steps with Python (solution)2a. First steps with Python and OpenCV (solution)

2b. Object detection (solution)

3. Tracking-by-Detection (solution)

4. First steps with TensorFlow (solution for linear regression task, solution for MLP task)

5. A CNN with TensorFlow (solution)

6. TensorBoard (solution)

7. Understanding CNNs by visualization (solution)

8. YOLO - You Only Look Once (solution)

9. YOLO9000: Better, Faster, Stronger (solution)

10. Experiments with YOLO

11. Single Shot MultiBox Detector (SSD) (solution)

12. Particle Filter

## Links

- Great visualization of a small CNN by Adam Harley
- Precision and Recall by Boris Babenko
- Understanding the mAP Evaluation Metric for Object Detection
- mAP (mean Average Precision) for Object Detection
- How to calculate mAP for detection task for the PASCAL VOC Challenge?
- Real-time Object Detection with YOLO, YOLOv2 and now YOLOv3 (article from March 2018)
- Understanding the Basis of the Kalman Filter
- Object detection models (from 42m:40s), Stanford Lecture 11 | Detection and Segmentation of CS 231n: Convolutional Neural Networks for Visual Recognition Excellent lecture by Justin Johnson.
- All lecture videos from 2017 version of Stanford Lecture "CS 231n: Convolutional Neural Networks for Visual Recognition"