Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google...
Transcript of Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google...
![Page 1: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/1.jpg)
Hacksession Image Recognition
Dr. Thorben Jensen
![Page 2: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/2.jpg)
Outline
07.11.2019
1 Intro Image Recognition
2 Object Detection and YOLO
3 Session Targets
4 Setup & Code Intro
2
![Page 3: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/3.jpg)
Intro Image Recognition
![Page 4: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/4.jpg)
Image classification in a nutshell
4Image source: https://www.mathworks.com
![Page 5: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/5.jpg)
Images are matrices
5
![Page 6: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/6.jpg)
Filters can match patterns
6
![Page 7: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/7.jpg)
Applying multiple filters
7
![Page 8: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/8.jpg)
To detect more complex features: apply filters after filters
8Image source: https://www.slideshare.net
Low-level Mid-level High-level Result
T. Cruise: 99 %
Input
T. Jensen: 1 %
![Page 9: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/9.jpg)
Object detection & YOLO
![Page 10: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/10.jpg)
Types of Image Recognition Tasks
07.11.2019 10
![Page 11: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/11.jpg)
YOLO – „you only look once“
Conventional Object Detection
separately proposes and classifies ‘boxes’
YOLO (“you only look once”)
parallelizes proposing and classifying ‘boxes’
07.11.2019 11
https://towardsdatascience.com/r-cnn-fast-r-cnn-faster-r-cnn-yolo-object-detection-algorithms-36d53571365e
![Page 12: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/12.jpg)
Network pre-trained on COCO-Dataset
• 80 object classes
• 330.000 images
• Networks pre-trained on COCO dataset freely available
07.11.2019INFORMATIONSFABRIK – Businesspräsentation 12
![Page 13: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/13.jpg)
Session Targets
![Page 14: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/14.jpg)
07.11.2019 15
https://56f2a99952126.streamlock.net/833/default.stream/playlist.m3u8
![Page 15: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/15.jpg)
1. Form groups (2-3 persons each, 1 Google account per group)
2. Intro into sample code
3. Diving into sample code with Google Colab
4. Choose a new use case, and code it with your group
– How many bicycles?
– Delivery trucks? (trucks only allowed at limited hours)
– Remove certain objects from an image?
– < your idea here >
5. 16:30-16:45: present your results to us
07.11.2019 16
Agenda
![Page 16: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/16.jpg)
Setup & Code Intro
![Page 17: Hacksession Image Recognition - c/o data science...1. Form groups (2-3 persons each, 1 Google account per group) 2. Intro into sample code 3. Diving into sample code with Google Colab](https://reader031.fdocuments.net/reader031/viewer/2022040908/5e803a3245010864963e584b/html5/thumbnails/17.jpg)
Contact
07.11.2019 18
Dr. Thorben Jensen Data Scientist
+49 160 69 666 42
www.informationsfabrik.de