Aktionen

Benutzerin

Verena: Unterschied zwischen den Versionen

Aus exmediawiki

Keine Bearbeitungszusammenfassung
 
(50 dazwischenliegende Versionen von einem anderen Benutzer werden nicht angezeigt)
Zeile 1: Zeile 1:
...
...


=KI-SEMINAR=
==Vorbereitungen==


==Projektdokumentation==
==== [[PHYTOTRON|PHYTOTRON – Workspace for Biological Media]] ====


===Kurzbeschreibung (EN)===
==== [[DIY_Computer|DIY Computer]] ====
 
==== [[Bewaesserungssystem|Bewässerungssystem für KHM-Garten]] ====
==== [[Workshop_Luftdaten|Workshop: Luftdaten]] ====
 
==== [[Rechnen durch Handeln]] ====
 
==== [[Mechatronics_101|Mechatronics 101]] ====
 
== Archiv ==
 
=== Workshop: [[DNA_phenotyping|Forensische DNA-Phänotypisierung]] ===
 
=== Seminar [[Open Lab]] ===
 
=== Seminar [[Whole_Earth_Reading_Group|WHOLE EARTH Reading Group]] ===
 
=== Ausstellungsprojekt [[Multispecies_@_Weltkunstzimmer|"GOOD BYE CRUEL WORLD, IT'S OVER"]] ===
 
=== Ausstellung [http://www.temporarygallery.org/?p=8780 Praktiken der Annäherung @ Temporary Gallery] ===
<!-- [[Practices_of_Approximation|Praktiken der Annäherung @ Temporary Gallery]] -->
 
=== Projekt [[KHM-Garten]] ===
=== Seminar [[Blockchain Reading Group|Blockchain Reading Group]] ===
=== Workshop [[KünstlerInnenhonorare]] ===
=== Seminar [[Re-Cycle?]] ===
 
<!--
==Linux==
===Bootfähigen USB-Stick erstellen (von Mac)===
 
 
Betriebssystem Ubuntu Desktop 18.04.3 LTS (last stable version) soll installiert werden
 
Anleitung: [https://tutorials.ubuntu.com/tutorial/tutorial-create-a-usb-stick-on-macos?_ga=2.75139794.1749488733.1571761439-720078860.1571761439#0 How to create a bootable USB stick on macOS]
 
2. Requirements
** A 2GB or larger USB stick/flash drive
** An Apple computer or laptop running macOS
** An Ubuntu ISO file. See [https://ubuntu.com/download/desktop Get Ubuntu] for download links
 
3. Prepare/Erase the USB stick Apple's ‘Disk Utility'
 
4. Install and run [https://www.balena.io/ Etcher]
 
5. Flash the USB drive
 
 
===Dual Boot auf Windows-System===
 
 
https://wiki.ubuntuusers.de/Dualboot/
 
<!--
==KI-SEMINAR==
 
===Projektdokumentation===
 
====Kurzbeschreibung (EN)====
<br>
<br>
'''THE OFFICE (working title), 2019'''
'''THE OFFICE (working title), 2019'''
Zeile 26: Zeile 84:
neural networks
neural networks


===Hintergrund/Research===
====Hintergrund/Research====




====Projekte Pflanzen & KI====
=====Projekte Pflanzen & KI=====


* [https://www.aestheticsofexclusion.com/botanicavariegata Botanica Variegata (2019?) von Sjoerd Ter Borg]
* [https://www.aestheticsofexclusion.com/projects/botanica-variegata Botanica Variegata (2019?) von Sjoerd Ter Borg]


====Projekte Environment & KI====
=====Projekte Environment & KI=====


* [http://harrischris.com/article/biophillic-vision-experiment-1 Biophillic Vision - Experiment 1: experimenting with machine learning to remove cars from video footage von Chris Harris]
* [http://harrischris.com/article/biophillic-vision-experiment-1 Biophillic Vision - Experiment 1: experimenting with machine learning to remove cars from video footage von Chris Harris]


===Technische Umsetzung/Vorgehensweise===
* https://rybakov.com/blog/ai_deleting_bodies/


====Erste Schritte====
====Technische Umsetzung/Vorgehensweise====
 
=====Erste Schritte=====


* Convolutional Neural Networks for object detection (in videos)
* Convolutional Neural Networks for object detection (in videos)
Zeile 48: Zeile 108:
* Heartbeat Tutorial Part 2: [https://heartbeat.fritz.ai/analyze-and-visualize-detected-video-objects-using-keras-and-imageai-d84c99b0ae8e Analyze and Visualize Detected Video Objects Using Keras and ImageAI]
* Heartbeat Tutorial Part 2: [https://heartbeat.fritz.ai/analyze-and-visualize-detected-video-objects-using-keras-and-imageai-d84c99b0ae8e Analyze and Visualize Detected Video Objects Using Keras and ImageAI]


====Aktueller Stand====
=====Aktueller Stand=====


Working with the pretrained network
Working with the pretrained network
Zeile 75: Zeile 135:
* write this data into a csv file
* write this data into a csv file


====Preparation of training, test and validation data====
=====Preparation of training, test and validation data=====




Zeile 123: Zeile 183:
* per batch processed image: output box points info of detected plant and, within each loop, crop image according to box points and save in newly created folder
* per batch processed image: output box points info of detected plant and, within each loop, crop image according to box points and save in newly created folder


====Nächste Schritte====
Cropped images final output:
* plant_center (ca. 2200 images)
* plant_chefofficefront(ca. 1300 images)
* plant_conferencefront (ca. 600 images)
* plant_conferenceinside (ca. 300 images)
* plant_receptiontop (ca. 738 images)
 
=====Nächste Schritte=====


* train custom model based on collected images
* train custom model based on collected images
Zeile 133: Zeile 200:
* video post production (if neccessary)
* video post production (if neccessary)


===Code===
====Code====


==Links zum Thema KI==
===Links zum Thema KI===


* [https://www.nature.com/articles/d41586-019-01413-1?fbclid=IwAR3DD0D66JN5qjxGuH8YsvNPxKLJs_NXuNzrLaopJ1jtFgGk3TNa_rtcSKs Don’t let industry write the rules for AI]
* [https://www.nature.com/articles/d41586-019-01413-1?fbclid=IwAR3DD0D66JN5qjxGuH8YsvNPxKLJs_NXuNzrLaopJ1jtFgGk3TNa_rtcSKs Don’t let industry write the rules for AI]
* [https://www.wired.com/story/ai-text-generator-too-dangerous-to-make-public/?utm_brand=wired&utm_source=facebook&utm_social-type=owned&utm_medium=social&mbid=social_fb&utm_campaign=wired&fbclid=IwAR0cVyOvfT4uDVeO66zjlPHOXbuS8aGlDChgKSPOd1jT_16YJ492BRxeOS4 The AI Text Generator That's Too Dangerous to Make Public]
* [https://www.wired.com/story/ai-text-generator-too-dangerous-to-make-public/?utm_brand=wired&utm_source=facebook&utm_social-type=owned&utm_medium=social&mbid=social_fb&utm_campaign=wired&fbclid=IwAR0cVyOvfT4uDVeO66zjlPHOXbuS8aGlDChgKSPOd1jT_16YJ492BRxeOS4 The AI Text Generator That's Too Dangerous to Make Public]


===Evolutionary Algorithms===
====Evolutionary Algorithms====
* [https://www.technologyreview.com/s/611568/evolutionary-algorithm-outperforms-deep-learning-machines-at-video-games/?utm_medium=social&utm_source=facebook.com&utm_campaign=owned_social&fbclid=IwAR01pdD_uNn4tobm2tpdDRibCaigDfiJgmUG9DW6GR5u_ltmdvfdAol4esQ Evolutionary algorithm outperforms deep-learning machines at video games], MIT Technology Review, 06/2018
* [https://www.technologyreview.com/s/611568/evolutionary-algorithm-outperforms-deep-learning-machines-at-video-games/?utm_medium=social&utm_source=facebook.com&utm_campaign=owned_social&fbclid=IwAR01pdD_uNn4tobm2tpdDRibCaigDfiJgmUG9DW6GR5u_ltmdvfdAol4esQ Evolutionary algorithm outperforms deep-learning machines at video games], MIT Technology Review, 06/2018


==Code==
===Code===




<!-- unsichtbarer Kommentar


=BLOCKCHAIN READING GROUP notes=
==Other Research==
 
===Teaching & Non-Teaching===
* [https://medium.com/pi-top/meet-the-school-with-no-classes-no-classrooms-and-no-curriculum-7cc7be517cef?fbclid=IwAR34KmSoA7ER59npOAu2PyE6DAQDZ5IthTloXfqmOD3rtMdOmaPm_LEpOek Meet the school with no classes, no classrooms and no curriculum], Medium, 05/2019
 
 
=BLOCKCHAIN READING GROUP notes=


> seminar page: [[Blockchain Reading Group|Blockchain Reading Group]]
> seminar page: [[Blockchain Reading Group|Blockchain Reading Group]]
Zeile 169: Zeile 243:
** https://en.wikipedia.org/wiki/Smart_contract
** https://en.wikipedia.org/wiki/Smart_contract
* [https://hackernoon.com/6-interesting-blockchain-projects-8c315364ff7f 6 Interesting Blockchain Projects]
* [https://hackernoon.com/6-interesting-blockchain-projects-8c315364ff7f 6 Interesting Blockchain Projects]
-->
-->


 
-->
=KünstlerInnenhonorare=
 
* Linksammlung zum Thema [[KünstlerInnenhonorare|KünstlerInnenhonorare]]
 
=Other Research=
 
==Teaching & Non-Teaching==
* [https://medium.com/pi-top/meet-the-school-with-no-classes-no-classrooms-and-no-curriculum-7cc7be517cef?fbclid=IwAR34KmSoA7ER59npOAu2PyE6DAQDZ5IthTloXfqmOD3rtMdOmaPm_LEpOek Meet the school with no classes, no classrooms and no curriculum], Medium, 05/2019

Aktuelle Version vom 11. Mai 2025, 17:03 Uhr