Aktionen

Benutzer

C.heck: Unterschied zwischen den Versionen

Aus exmediawiki
 
(99 dazwischenliegende Versionen desselben Benutzers werden nicht angezeigt)
Zeile 1: Zeile 1:
 +
<big>'''test- bzw. researchpages...:
 +
'''
 +
</big>
 +
{| class="wikitable"
 +
|-
 +
| [[Maschinelles Sprechen]] ||[[Per noMachine auf AI-Lab-Rechner zugreifen]] || [[Maschinelles Lernen]] || [[Maschinelles Lesen]] || [[Maschinelles Dichten]]
 +
|-
 +
| [[KHM-Wolke einrichten]] || [["...Sprache"]] || [[Vergleichsoperatoren]] || [[Perzeptron]] || [[Boole]]
 +
|-
 +
| [[Code Poetry]] || [[tesssts]] || [[AI@exLabIII]] || [[Einführung in die Programmierung Künstlicher Intelligenzen]] || [[Future Minds – Kritik Künstlicher Intelligenzen]]
 +
|-
 +
| [[Kommandozeilenreferenz]] || [[jutta-weber-kommentar]] || [[Die Kommandozeile]] || [[praktische Erinnerungskultur]] || [[Krieg & KI]]
 +
|-
 +
| [[das Codicht]] || [[machtkaputtwaseuchkaputtmachtstuktur]] || [https://pad.freifunk.net/p/W74NqM6ekDu01AsMK4Qu PAD-exp-inf.-talk@aula] || [https://pad.freifunk.net/p/newlinkski PAD-ki_research]|| [https://pad.freifunk.net/p/sdV7m6aeNV_-87gPO3qF PAD-booklist]
 +
|-
 +
| [https://pad.freifunk.net/p/d8jqQlhq2wzpPN5BFL-O PAD-wo-beginnt-der-krieg_abstract] || [[notes_rechnen mit lebenden systeme]] || [https://pad.freifunk.net/p/J63t2lAnTCC5yjjIoHYR seminar-notes] || [[AI on Rasp Pi]] || [[krieg-u-ki-test]]
 +
|-
 +
| [[Per Teamviewer auf AI-Lab-Rechner zugreifen]] || [[Kriegseinführung]] || [[Krieg-u-Ki-Archivseite]] || [[notes-poiesis_SS18]] || [[Poiesis and Pottery]]
 +
|-
 +
| [[RunwayML]] || [[ChatterBot]] || [[python-intro-brainstorming]] || [[TextBlob]] || [[sentiment analysis]]
 +
|}
  
 +
[[KI-Newsletter 4]]
 +
 +
---
 +
[[how to get your trainigdata]]
 +
----
 +
 +
esolangs:
 +
* [[Beatnik]]
 +
 +
----
 +
 +
[[Hacks]]
 +
 +
----
 +
 +
jupyter:
 +
[[jupyter-extensions]]
 +
 +
[[KHM-Cloud]]
 +
 +
[[AI Dungeon]]
 +
 +
----
 +
digitale poesie:
 +
 +
https://0x0a.li/de/
 +
https://poesiefestival.org/de/mediathek/digitale-poesie/
 +
 +
dann die arbeiten von thurston, bajohr, morris, goldsmith etc...
 +
 +
https://de.wikipedia.org/wiki/Digitale_Poesie
 +
 +
----
 +
 +
[[testpage-pdfhandler]]
 +
 +
[[exmediawiki-editing]]
 +
 +
[[another-testpage]]
 +
 +
[[Big Blue Button]]
 +
 +
=Keras Examples=
 +
https://github.com/keras-team/keras/tree/master/examples
 +
----
 
einfaches perceptron (schöne skizzen): https://github.com/nature-of-code/NOC-S17-2-Intelligence-Learning/blob/master/week4-neural-networks/perceptron.pdf  
 
einfaches perceptron (schöne skizzen): https://github.com/nature-of-code/NOC-S17-2-Intelligence-Learning/blob/master/week4-neural-networks/perceptron.pdf  
 +
 +
----
 +
----
 +
 
=ADVERSARIAL ATTACKS=
 
=ADVERSARIAL ATTACKS=
 
KNN's sind extrem anfällig für...
 
KNN's sind extrem anfällig für...
−
 
 
* Praxis-Beispiele: https://boingboing.net/tag/adversarial-examples
 
* Praxis-Beispiele: https://boingboing.net/tag/adversarial-examples
 
* https://bdtechtalks.com/2018/12/27/deep-learning-adversarial-attacks-ai-malware/
 
* https://bdtechtalks.com/2018/12/27/deep-learning-adversarial-attacks-ai-malware/
Zeile 11: Zeile 80:
  
 
* https://en.wikipedia.org/wiki/Deep_learning#Cyberthreat
 
* https://en.wikipedia.org/wiki/Deep_learning#Cyberthreat
−
+
 
 +
----
 +
----
 
=WHITE BOX ATTACKS=
 
=WHITE BOX ATTACKS=
 
* https://cv-tricks.com/how-to/breaking-deep-learning-with-adversarial-examples-using-tensorflow/
 
* https://cv-tricks.com/how-to/breaking-deep-learning-with-adversarial-examples-using-tensorflow/
 
** Paper »ADVERSARIAL EXAMPLES IN THE PHYSICAL WORLD«: https://arxiv.org/pdf/1607.02533.pdf
 
** Paper »ADVERSARIAL EXAMPLES IN THE PHYSICAL WORLD«: https://arxiv.org/pdf/1607.02533.pdf
  
 +
----
 
==Untargeted Adversarial Attacks==
 
==Untargeted Adversarial Attacks==
 
Adversarial attacks that just want '''your model to be confused and predict a wrong class''' are called Untargeted Adversarial Attacks.
 
Adversarial attacks that just want '''your model to be confused and predict a wrong class''' are called Untargeted Adversarial Attacks.
 
* nicht zielgerichtet
 
* nicht zielgerichtet
 +
 
===Fast Gradient Sign Method(FGSM)===
 
===Fast Gradient Sign Method(FGSM)===
 
FGSM is a single step attack, ie.. the perturbation is added in a single step instead of adding it over a loop (Iterative attack).
 
FGSM is a single step attack, ie.. the perturbation is added in a single step instead of adding it over a loop (Iterative attack).
 +
 
===Basic Iterative Method===
 
===Basic Iterative Method===
 
Störung, anstatt in einem einzelnen Schritt in mehrere kleinen Schrittgrößen anwenden
 
Störung, anstatt in einem einzelnen Schritt in mehrere kleinen Schrittgrößen anwenden
 +
 
===Iterative Least-Likely Class Method===
 
===Iterative Least-Likely Class Method===
 
ein Bild erstellen, welches in der Vorhersage den niedrigsten Score trägt
 
ein Bild erstellen, welches in der Vorhersage den niedrigsten Score trägt
 +
 +
----
 
==Targeted Adversarial Attacks==
 
==Targeted Adversarial Attacks==
 
Attacks which compel the model to predict a '''(wrong) desired output''' are called Targeted Adversarial attacks
 
Attacks which compel the model to predict a '''(wrong) desired output''' are called Targeted Adversarial attacks
 
* zielgerichtet
 
* zielgerichtet
 +
 +
----
 
==(Un-)Targeted Adversarial Attacks==
 
==(Un-)Targeted Adversarial Attacks==
 
kann beides...
 
kann beides...
 +
 
===Projected Gradient Descent (PGD)===
 
===Projected Gradient Descent (PGD)===
 
Eine Störung finden die den Verlust eines Modells bei einer bestimmten Eingabe maximiert:
 
Eine Störung finden die den Verlust eines Modells bei einer bestimmten Eingabe maximiert:
Zeile 35: Zeile 115:
 
** Jupyter Notebook: https://github.com/oscarknagg/adversarial/blob/master/notebooks/Creating_And_Defending_From_Adversarial_Examples.ipynb
 
** Jupyter Notebook: https://github.com/oscarknagg/adversarial/blob/master/notebooks/Creating_And_Defending_From_Adversarial_Examples.ipynb
  
 +
----
 +
----
 
=BLACK BOX ATTACKS=
 
=BLACK BOX ATTACKS=
  
Zeile 40: Zeile 122:
 
** Jupyter Notebook: https://github.com/dangeng/Simple_Adversarial_Examples
 
** Jupyter Notebook: https://github.com/dangeng/Simple_Adversarial_Examples
  
 +
----
 
==on computer vision==
 
==on computer vision==
 +
 
===propose zeroth order optimization (ZOO)===
 
===propose zeroth order optimization (ZOO)===
 
* attacks to directly estimate the gradients of the targeted DNN
 
* attacks to directly estimate the gradients of the targeted DNN
 
** https://arxiv.org/abs/1708.03999
 
** https://arxiv.org/abs/1708.03999
 +
 
===Black-Box Attacks using Adversarial Samples===
 
===Black-Box Attacks using Adversarial Samples===
 
*  a technique that uses the victim model as an oracle to label a synthetic training set for the substitute, so the attacker need not even collect a training set to mount the attack
 
*  a technique that uses the victim model as an oracle to label a synthetic training set for the substitute, so the attacker need not even collect a training set to mount the attack
 
** https://arxiv.org/abs/1605.07277
 
** https://arxiv.org/abs/1605.07277
 +
 
===new Tesla Hack===
 
===new Tesla Hack===
 
* https://spectrum.ieee.org/cars-that-think/transportation/self-driving/three-small-stickers-on-road-can-steer-tesla-autopilot-into-oncoming-lane
 
* https://spectrum.ieee.org/cars-that-think/transportation/self-driving/three-small-stickers-on-road-can-steer-tesla-autopilot-into-oncoming-lane
Zeile 52: Zeile 138:
 
** Paper vom Forschungsteam: https://keenlab.tencent.com/en/whitepapers/Experimental_Security_Research_of_Tesla_Autopilot.pdf
 
** Paper vom Forschungsteam: https://keenlab.tencent.com/en/whitepapers/Experimental_Security_Research_of_Tesla_Autopilot.pdf
  
 +
----
 
==on voice (ASR)==
 
==on voice (ASR)==
 
* https://www.the-ambient.com/features/weird-ways-echo-can-be-hacked-how-to-stop-it-231
 
* https://www.the-ambient.com/features/weird-ways-echo-can-be-hacked-how-to-stop-it-231
 +
 
===hidden voice commands===
 
===hidden voice commands===
 
* https://www.theregister.co.uk/2016/07/11/siri_hacking_phones/
 
* https://www.theregister.co.uk/2016/07/11/siri_hacking_phones/
 
* https://www.fastcompany.com/90240975/alexa-can-be-hacked-by-chirping-birds
 
* https://www.fastcompany.com/90240975/alexa-can-be-hacked-by-chirping-birds
 +
 
===Psychoacoustic Hiding (Attacking Speech Recognition)===
 
===Psychoacoustic Hiding (Attacking Speech Recognition)===
 
* https://adversarial-attacks.net/
 
* https://adversarial-attacks.net/
Zeile 63: Zeile 152:
 
** Präsentationsfolien: https://www.ndss-symposium.org/wp-content/uploads/ndss2019_08-2_Schonherr_slides.pdf
 
** Präsentationsfolien: https://www.ndss-symposium.org/wp-content/uploads/ndss2019_08-2_Schonherr_slides.pdf
  
 +
----
 
==on written text (NLP)==
 
==on written text (NLP)==
 +
 
===paraphrasing attacks===
 
===paraphrasing attacks===
 
* https://venturebeat.com/2019/04/01/text-based-ai-models-are-vulnerable-to-paraphrasing-attacks-researchers-find/
 
* https://venturebeat.com/2019/04/01/text-based-ai-models-are-vulnerable-to-paraphrasing-attacks-researchers-find/
Zeile 70: Zeile 161:
 
* https://motherboard.vice.com/en_us/article/9axx5e/ai-can-be-fooled-with-one-misspelled-word
 
* https://motherboard.vice.com/en_us/article/9axx5e/ai-can-be-fooled-with-one-misspelled-word
  
 +
----
 
==Anti Surveillance==
 
==Anti Surveillance==
 
http://dismagazine.com/dystopia/evolved-lifestyles/8115/anti-surveillance-how-to-hide-from-machines/
 
http://dismagazine.com/dystopia/evolved-lifestyles/8115/anti-surveillance-how-to-hide-from-machines/
  
 +
----
 
==libraries==
 
==libraries==
 
* https://github.com/bethgelab
 
* https://github.com/bethgelab
 
* https://github.com/tensorflow/cleverhans
 
* https://github.com/tensorflow/cleverhans
  
−
=ETHIK=
+
----
 +
----
 +
=ETHICS=
 
* https://www.economist.com/science-and-technology/2018/02/15/computer-programs-recognise-white-men-better-than-black-women
 
* https://www.economist.com/science-and-technology/2018/02/15/computer-programs-recognise-white-men-better-than-black-women
 
* https://books.google.de/books?id=rLsyDwAAQBAJ&pg=PA95&redir_esc=y#v=onepage&q&f=false
 
* https://books.google.de/books?id=rLsyDwAAQBAJ&pg=PA95&redir_esc=y#v=onepage&q&f=false
Zeile 85: Zeile 180:
 
* AI Now Report: https://medium.com/@AINowInstitute/the-10-top-recommendations-for-the-ai-field-in-2017-b3253624a7
 
* AI Now Report: https://medium.com/@AINowInstitute/the-10-top-recommendations-for-the-ai-field-in-2017-b3253624a7
 
* https://bdtechtalks.com/2018/03/26/racist-sexist-ai-deep-learning-algorithms/
 
* https://bdtechtalks.com/2018/03/26/racist-sexist-ai-deep-learning-algorithms/
 +
* https://www.spiegel.de/netzwelt/web/facebook-foerdert-die-ki-forschung-an-der-tu-muenchen-gastbeitrag-a-1250796.html#ref=rss
  
 +
----
 +
----
 
=XAI=
 
=XAI=
 
[[XAI/NLG]]
 
[[XAI/NLG]]
Zeile 97: Zeile 195:
 
* DARPA: https://www.darpa.mil/program/explainable-artificial-intelligence
 
* DARPA: https://www.darpa.mil/program/explainable-artificial-intelligence
  
 +
 +
Bedeutung:
 +
* https://de.m.wikipedia.org/wiki/Bedeutung_(Sprachphilosophie)
 +
* https://de.m.wikipedia.org/wiki/Philosophie_des_Geistes
 +
* https://de.m.wikipedia.org/wiki/Mustererkennung
 +
 +
 +
----
 +
==XAI durch Sprachrationalisierung==
 +
* https://de.wikipedia.org/wiki/Rationalismus
 +
** Baumgarten ... << Rationalismus und Ästhetik
 +
*** sinnliche Erfahrung v.s. Wissen a priori << wo stehen wir hier mit der ästhetischen Erfahrung?
 +
*** https://www.kubi-online.de/artikel/aesthetische-erfahrung
 +
 +
* Rationalization: A Neural Machine Translation Approach to Generating Natural Language Explanations
 +
** https://arxiv.org/abs/1702.07826
 +
 +
----
 +
----
 
=LANGUAGE=
 
=LANGUAGE=
−
==esotheric neural net (programming language)==
+
 
 +
----
 +
==Konstruierte, bzw. Künstliche Sprachen==
 +
* https://www.fatum-magazin.de/ausgaben/dialog/praefrontal/kunstsprachen.html
 +
* https://de.wikipedia.org/wiki/Konstruierte_Sprache
 +
* http://deacademic.com/dic.nsf/dewiki/441004
 +
** https://de.m.wikipedia.org/wiki/Neusprech
 +
** https://de.m.wikipedia.org/wiki/Weltdeutsch
 +
 
 +
===esoterische (KI) Programmiersprachen===
 +
* https://esolangs.org/w/index.php?search=neural&title=Special%3ASearch&profile=default&fulltext=1
 +
* esoterische programmiersprachen http://kryptografie.de/kryptografie/chiffre/index-sprachen.htm
 +
 
 
* Forscher suchen eigene Programmiersprache: https://t3n.de/news/machine-learning-facebooks-ki-chef-sucht-sprache-1144900/
 
* Forscher suchen eigene Programmiersprache: https://t3n.de/news/machine-learning-facebooks-ki-chef-sucht-sprache-1144900/
−
* esoterische programmiersprachen http://kryptografie.de/kryptografie/chiffre/index-sprachen.htm
 
  
 +
----
 +
==Computer (und) Literatur==
 +
* p0es1s: http://www.p0es1s.net/p0es1s/intro_d.htm
 +
* stochastische Texte: https://auer.netzliteratur.net/0_lutz/lutz_original.html
 +
* Netzliteratur (Projekte): https://netzliteratur.net/netzliteratur_projekte_a.php
 +
 +
Florian Cramer:
 +
* https://www.netzliteratur.net/cramer/poetische_kalkuele_und_phantasmen.pdf
 +
* https://www.netzliteratur.net/cramer/wordsmadefleshpdf.pdf
 +
 +
----
 +
==KI und Literatur==
 +
https://www.faz.net/aktuell/feuilleton/buecher/literatur-und-ki-vernunft-ist-auch-eine-herzenssache-16079038.html?printPagedArticle=true#void
 +
 +
===AI-Poetry Examples...===
 +
* https://hackernoon.com/i-tried-my-hand-at-deep-learning-and-made-some-poetry-along-the-way-2e350c33376f
 +
* https://www.japandigest.de/aktuelles/technologie-roboter/kunstliche-intelligenz-schreibt-haiku/
 +
* https://bgr.com/2018/08/08/poetry-ai-bot-shakespeare-human-research/
 +
* https://www.konstantext.de/index.php/item/sonnenblicke-auf-der-flucht-wenn-kuenstliche-intelligenz-ein-gedicht-schreibt
 +
 +
----
 +
==(KI-generierte) Krypto==
 +
* https://motherboard.vice.com/de/article/8q8wkv/google-ki-entwickelt-verschluesselung-die-selbst-google-nicht-versteht
 +
* http://kryptografie.de/kryptografie/index.htm
 +
 +
----
 
==NLU / NLI==
 
==NLU / NLI==
 
* https://www.informatik-aktuell.de/betrieb/kuenstliche-intelligenz/natural-language-understanding-nlu.html
 
* https://www.informatik-aktuell.de/betrieb/kuenstliche-intelligenz/natural-language-understanding-nlu.html
 
* https://en.wikipedia.org/wiki/Natural-language_understanding
 
* https://en.wikipedia.org/wiki/Natural-language_understanding
  
 +
----
 
==NLP==
 
==NLP==
 
* https://de.wikipedia.org/wiki/Computerlinguistik
 
* https://de.wikipedia.org/wiki/Computerlinguistik
  
 
===Speech recognition===
 
===Speech recognition===
−
https://de.wikipedia.org/wiki/Spracherkennung
+
* https://de.wikipedia.org/wiki/Spracherkennung
 +
* https://www.golem.de/news/deep-speech-0-2-mozillas-spracherkennung-wird-kleiner-und-kann-echtzeit-1809-136645.html
 +
** Code: https://github.com/mozilla/DeepSpeech/releases/tag/v0.2.0
 +
 
 +
----
  
 
==NLG==
 
==NLG==
 +
https://byteacademy.co/blog/overview-NLG
 +
 
* https://de.wikipedia.org/wiki/Textgenerierung
 
* https://de.wikipedia.org/wiki/Textgenerierung
 
* http://www.thealit.de/lab/serialitaet/teil/nieberle/nieberle.html
 
* http://www.thealit.de/lab/serialitaet/teil/nieberle/nieberle.html
Zeile 125: Zeile 286:
 
** https://www.fastcompany.com/90132632/ai-is-inventing-its-own-perfect-languages-should-we-let-it
 
** https://www.fastcompany.com/90132632/ai-is-inventing-its-own-perfect-languages-should-we-let-it
  
−
===techniques===
+
----
−
====LSTM+RNN===
+
===(un-)supervised techniques===
 +
 
 +
====LSTM====
 +
http://colah.github.io/posts/2015-08-Understanding-LSTMs/
 +
 
 +
====LSTM+RNN====
 
* »on the road« by AI: https://medium.com/artists-and-machine-intelligence/ai-poetry-hits-the-road-eb685dfc1544
 
* »on the road« by AI: https://medium.com/artists-and-machine-intelligence/ai-poetry-hits-the-road-eb685dfc1544
−
====Autoencoders====
+
 
 +
====Autoencoder====
 
https://www.wired.co.uk/article/google-artificial-intelligence-poetry
 
https://www.wired.co.uk/article/google-artificial-intelligence-poetry
 +
 +
====LSTM+Autoencoder====
 +
* https://github.com/keras-team/keras/issues/1401
 +
* https://www.dlology.com/blog/how-to-do-unsupervised-clustering-with-keras/
 +
 
====GAN====
 
====GAN====
−
====Transformations====
+
https://arxiv.org/abs/1705.10929
 +
 
 +
====transformer-based language model====
 +
OpenAI's gpt-2:
 +
* https://openai.com/blog/better-language-models/
 +
** https://github.com/openai/gpt-2
  
−
===Poetry===
+
Diskussion:
−
====examples...====
+
* https://www.skynettoday.com/briefs/gpt2
−
* https://hackernoon.com/i-tried-my-hand-at-deep-learning-and-made-some-poetry-along-the-way-2e350c33376f
 
−
* https://www.japandigest.de/aktuelles/technologie-roboter/kunstliche-intelligenz-schreibt-haiku/
 
−
* https://bgr.com/2018/08/08/poetry-ai-bot-shakespeare-human-research/
 
  
 +
----
 
===datenbanken===
 
===datenbanken===
 
deutsch:
 
deutsch:
Zeile 150: Zeile 325:
 
* http://www.macs.hw.ac.uk/InteractionLab/E2E/
 
* http://www.macs.hw.ac.uk/InteractionLab/E2E/
  
 +
----
 
===chatbots===
 
===chatbots===
 
* https://bdtechtalks.com/2017/08/21/rob-high-ibm-watson-cto-artificial-intelligence-chatbots/
 
* https://bdtechtalks.com/2017/08/21/rob-high-ibm-watson-cto-artificial-intelligence-chatbots/
Zeile 159: Zeile 335:
 
** Jupyter Notebooks: https://github.com/suriyadeepan/practical_seq2seq
 
** Jupyter Notebooks: https://github.com/suriyadeepan/practical_seq2seq
  
 +
----
 
===Toolkits/Librarys===
 
===Toolkits/Librarys===
 
* Natural Language Toolkit: http://www.nltk.org/
 
* Natural Language Toolkit: http://www.nltk.org/
Zeile 167: Zeile 344:
 
* https://github.com/shashank-bhatt-07/Natural-Language-Generation-using-LSTM-Keras
 
* https://github.com/shashank-bhatt-07/Natural-Language-Generation-using-LSTM-Keras
  
−
=KI-GENERIERTE KRYPTO=
+
----
−
* https://motherboard.vice.com/de/article/8q8wkv/google-ki-entwickelt-verschluesselung-die-selbst-google-nicht-versteht
+
----
−
* http://kryptografie.de/kryptografie/index.htm
 
−
 
 
 
=REPRODUKTIVE KI=
 
=REPRODUKTIVE KI=
 
https://www.sir-apfelot.de/kuenstliche-intelligenz-erschafft-neue-ki-systeme-10436/
 
https://www.sir-apfelot.de/kuenstliche-intelligenz-erschafft-neue-ki-systeme-10436/

Aktuelle Version vom 19. Februar 2021, 22:14 Uhr

test- bzw. researchpages...:

Maschinelles Sprechen Per noMachine auf AI-Lab-Rechner zugreifen Maschinelles Lernen Maschinelles Lesen Maschinelles Dichten
KHM-Wolke einrichten "...Sprache" Vergleichsoperatoren Perzeptron Boole
Code Poetry tesssts AI@exLabIII Einführung in die Programmierung Künstlicher Intelligenzen Future Minds – Kritik Künstlicher Intelligenzen
Kommandozeilenreferenz jutta-weber-kommentar Die Kommandozeile praktische Erinnerungskultur Krieg & KI
das Codicht machtkaputtwaseuchkaputtmachtstuktur PAD-exp-inf.-talk@aula PAD-ki_research PAD-booklist
PAD-wo-beginnt-der-krieg_abstract notes_rechnen mit lebenden systeme seminar-notes AI on Rasp Pi krieg-u-ki-test
Per Teamviewer auf AI-Lab-Rechner zugreifen Kriegseinführung Krieg-u-Ki-Archivseite notes-poiesis_SS18 Poiesis and Pottery
RunwayML ChatterBot python-intro-brainstorming TextBlob sentiment analysis

KI-Newsletter 4

--- how to get your trainigdata

esolangs:

Hacks

jupyter: jupyter-extensions

KHM-Cloud

AI Dungeon

digitale poesie:

https://0x0a.li/de/ https://poesiefestival.org/de/mediathek/digitale-poesie/

dann die arbeiten von thurston, bajohr, morris, goldsmith etc...

https://de.wikipedia.org/wiki/Digitale_Poesie

testpage-pdfhandler

exmediawiki-editing

another-testpage

Big Blue Button

Keras Examples

https://github.com/keras-team/keras/tree/master/examples

einfaches perceptron (schöne skizzen): https://github.com/nature-of-code/NOC-S17-2-Intelligence-Learning/blob/master/week4-neural-networks/perceptron.pdf

ADVERSARIAL ATTACKS

KNN's sind extrem anfällig für...

WHITE BOX ATTACKS

Untargeted Adversarial Attacks

Adversarial attacks that just want your model to be confused and predict a wrong class are called Untargeted Adversarial Attacks.

  • nicht zielgerichtet

Fast Gradient Sign Method(FGSM)

FGSM is a single step attack, ie.. the perturbation is added in a single step instead of adding it over a loop (Iterative attack).

Basic Iterative Method

Störung, anstatt in einem einzelnen Schritt in mehrere kleinen Schrittgrößen anwenden

Iterative Least-Likely Class Method

ein Bild erstellen, welches in der Vorhersage den niedrigsten Score trägt

Targeted Adversarial Attacks

Attacks which compel the model to predict a (wrong) desired output are called Targeted Adversarial attacks

  • zielgerichtet

(Un-)Targeted Adversarial Attacks

kann beides...

Projected Gradient Descent (PGD)

Eine Störung finden die den Verlust eines Modells bei einer bestimmten Eingabe maximiert:

BLACK BOX ATTACKS

on computer vision

propose zeroth order optimization (ZOO)

Black-Box Attacks using Adversarial Samples

  • a technique that uses the victim model as an oracle to label a synthetic training set for the substitute, so the attacker need not even collect a training set to mount the attack

new Tesla Hack

on voice (ASR)

hidden voice commands

Psychoacoustic Hiding (Attacking Speech Recognition)

on written text (NLP)

paraphrasing attacks

Anti Surveillance

http://dismagazine.com/dystopia/evolved-lifestyles/8115/anti-surveillance-how-to-hide-from-machines/

libraries

ETHICS

XAI

XAI/NLG


Bedeutung:


XAI durch Sprachrationalisierung

LANGUAGE

Konstruierte, bzw. Künstliche Sprachen

esoterische (KI) Programmiersprachen

Computer (und) Literatur

Florian Cramer:

KI und Literatur

https://www.faz.net/aktuell/feuilleton/buecher/literatur-und-ki-vernunft-ist-auch-eine-herzenssache-16079038.html?printPagedArticle=true#void

AI-Poetry Examples...

(KI-generierte) Krypto

NLU / NLI

NLP

Speech recognition

NLG

https://byteacademy.co/blog/overview-NLG

(un-)supervised techniques

LSTM

http://colah.github.io/posts/2015-08-Understanding-LSTMs/

LSTM+RNN

Autoencoder

https://www.wired.co.uk/article/google-artificial-intelligence-poetry

LSTM+Autoencoder

GAN

https://arxiv.org/abs/1705.10929

transformer-based language model

OpenAI's gpt-2:

Diskussion:

datenbanken

deutsch:

englisch:

E2E NLG Challenge:

chatbots

Toolkits/Librarys

tryouts:

REPRODUKTIVE KI

https://www.sir-apfelot.de/kuenstliche-intelligenz-erschafft-neue-ki-systeme-10436/

Abgerufen von „https://exmediawiki.khm.de/index.php?title=Benutzer:C.heck&oldid=6894“