Published May 14, 2020

Disrupting DeepFakes: Adversarial Attacks Against Conditional Image Translation Networks with...

Nataniel Ruiz delves into adversarial attacks to safeguard against deepfakes, examining methods to inject noise into image translation networks and highlighting the ethical and security challenges posed by generative models.

Episode Highlights

Topics covered

Episode Highlights

Related Episodes

Attacking Malware with Adversarial Machine Learning, w/ Edward Raff - #529
Answers 383 questions
Creative Adversarial Networks for Art Generation with Ahmed Elgammal - TWiML Talk #265
Answers 383 questions
Advancing Autonomous Vehicle Development Using Distributed Deep Learning with Adrien Gaidon -...
Answers 383 questions
Adversarial Attacks Against Reinforcement Learning Agents with Ian Goodfellow & Sandy Huang - #119
Answers 383 questions
Privacy and Security for Stable Diffusion and LLMs with Nicholas Carlini - 618
Answers 383 questions
Deep Learning, Transformers, and the Consequences of Scale with Oriol Vinyals - #546
Answers 383 questions
Trends in Machine Learning with Anima Anandkumar - TWiML Talk #215
Answers 383 questions
Towards Improved Transfer Learning with Hugo Larochelle - 631
Answers 383 questions
Deep Learning is Eating 5G. Here’s How, w/ Joseph Soriaga - #525
Answers 383 questions
Coercing LLMs to Do and Reveal (Almost) Anything with Jonas Geiping - 678
Answers 383 questions
Watermarking LLMs to Fight Plagiarism with Tom Goldstein - 621
Answers 383 questions
Stealing Weights of a Production LLM Like OpenAI’s ChatGPT with Nicholas Carlini - 702
Answers 383 questions
Dissecting the Controversy around OpenAI's New Language Model
Answers 383 questions
Deep Learning: Modular in Theory, Inflexible in Practice with Diogo Almeida - #8
Answers 383 questions
Offensive vs Defensive Data Science with Deep Varma - #25
Answers 383 questions

Dexa/This Week in ML & AI

Disrupting DeepFakes: Adversarial Attacks Against Conditional Image Translation Networks with...

Topics covered

Popular Clips

Stay Inside Safely

Show Wrap-Up

Staying Safe

Exploring AI Innovations

Understanding Adversarial Attacks

Image Manipulation Defense

Understanding Adversarial Attacks

AI Insights Unpacked

Attacking Image Translation

Generative Models and Ethics

Closing Thoughts

Key Contributions Unpacked

Adversarial Training Insights

Closing Thoughts

Episode Highlights

Image Translation Networks

Generative Models

Deepfake Disruption

Related Episodes

Attacking Malware with Adversarial Machine Learning, w/ Edward Raff - #529

Creative Adversarial Networks for Art Generation with Ahmed Elgammal - TWiML Talk #265

Advancing Autonomous Vehicle Development Using Distributed Deep Learning with Adrien Gaidon -...

Adversarial Attacks Against Reinforcement Learning Agents with Ian Goodfellow & Sandy Huang - #119

Privacy and Security for Stable Diffusion and LLMs with Nicholas Carlini - 618

Deep Learning, Transformers, and the Consequences of Scale with Oriol Vinyals - #546

Trends in Machine Learning with Anima Anandkumar - TWiML Talk #215

Towards Improved Transfer Learning with Hugo Larochelle - 631

Deep Learning is Eating 5G. Here’s How, w/ Joseph Soriaga - #525

Coercing LLMs to Do and Reveal (Almost) Anything with Jonas Geiping - 678

Watermarking LLMs to Fight Plagiarism with Tom Goldstein - 621

Stealing Weights of a Production LLM Like OpenAI’s ChatGPT with Nicholas Carlini - 702

Dissecting the Controversy around OpenAI's New Language Model

Deep Learning: Modular in Theory, Inflexible in Practice with Diogo Almeida - #8

Offensive vs Defensive Data Science with Deep Varma - #25

Disrupting DeepFakes: Adversarial Attacks Against Conditional Image Translation Networks with...

Topics covered

Popular Clips

Episode Highlights

Image Translation Networks

Generative ModelsNataniel Ruiz explores the ethical and technical challenges of generative models, particularly in the context of deepfakes. He discusses adversarial attacks as a method to protect against unauthorized image manipulation.

Generative Models

Deepfake DisruptionNataniel Ruiz explores adversarial attacks on deepfake systems, focusing on techniques to disrupt image translation networks. He discusses vulnerabilities in these systems and the role of noise in enhancing attack strategies.

Deepfake Disruption

Related Episodes