Image Editing from Text Instructions with AI!

Image Editing from Text Instructions with AI!

Good morning fellow AI enthusiast! This week's iteration focuses on a new image editing model called InstructPix2Pix, which, as its name says, translates a picture into another one following text instructions. We also welcome our new ethicist, Auxane, this iteration commenting on generative AI! We hope you enjoy it.

Receive the weekly digest right in your emails ??

1?? Image Editing from Text Instructions with AI!

No alt text provided for this image
Example from the InstructPix2Pix paper.

We know that AI can generate images; now, let’s edit them!

This new model called InstructPix2Pix does precisely that; it edits an image following a text-based instruction given by the user. Just look at those amazing results… and that is not from OpenAI or google with an infinite budget.

It is a recent publication from Tim Brooks and collaborators at the University of California, including prof. Alexei A. Efros, a well-known figure in the computer vision industry. As you can see, the results are just incredible.

Feel free to try it yourself, as they shared a demo of the model on huggingface, which you can use for free!

What’s even more incredible is how they achieved those results. Let’s dive into it! (or learn more in the article)


Join the discussion with fellow AI enthusiasts ??

2?? AI Ethics with Auxane Boch

No alt text provided for this image

But first, let's introduce Auxane!

Auxane is a researcher at?iuvenal research, a?Doctoral Candidate at the School of Management of the Technical University of Munich (TUM), and a research associate?at the TUM Institute for Ethics in Artificial Intelligence (IEAI).?

What a resume!

After studying psychology in France (B.Sc.), and cyberpsychology in Ireland (M.Sc.), her current research focuses on accountability for AI systems, social robotics, video games and digital education. Auxane also?works closely with the Women in AI group on the educational side, and is a Women in Games ambassador.

She is excited to join us for the ethics segment of those newsletters, and is always happy to chit-chat if you want to reach out. You can contact her on her linkedin, or use her email address ([email protected])!

Now for Auxane's ethics bite...

The use of AI to modify images brings with it a lot of exciting fantastic perspectives, but also?a lot of legal and ethical questions. Let’s dive into the major points to keep in mind!?If we look closer into the use of AI for digital arts, two school of arguments face each other; the first argues that the use of AI will impede creativity, while the other assert?that AI is only but a tool and facilitator. The truth, as usual, probably lies in the middle. But, when considering digital arts and creations, another question arises: what about the legal aspect? If AI is used to create a piece, does it legally belong to the AI, to the entity who created the AI,?or to the human who used the AI? Those questions are not clearly answered yet, and arguments could be made from each perspective. Finally, I think it’s important to see the usefulness of such tools as a strong positive point to support the evolution of the technology. Yes, it can be used for deep fakes and sometimes maleficent purposes, but with the right frame of use, it can also support architects or landscapers to project what a building or a city will look like with their proposed work, or?by historians and teacher to generate accurate and detailed reproductions of old cities. As per usual in technology and ethics, the problem is rarely the tool, but the way we use it.?

Segment by Auxane Boch


We are extremely grateful that?the newsletter?is now read by over 11'000+ incredible human beings counting our email list and LinkedIn subscribers. Feel free to reach out [email protected]?with any questions or details on sponsorships. Feel free to follow our newsletter at Towards AI, sharing the most exciting news, learning resources, articles, and memes from our Discord community weekly.

Thank you for reading and we wish you a fantastic week! Be sure to have?enough rest and sleep!

We will see you next week with another amazing paper!

Louis


Louis-Fran?ois Bouchard

Making AI accessible. ?? What's AI on YouTube. Co-founder at Towards AI. ex-PhD Student.

1 年

Ethics segment by Auxane Boch :)

Mike Nash (BA HONS)

Generative AI | AI | business growth finder and advisor. Get the most from AI with minimal risk - AI strategy, AI insights and leading AI advice - Contact me today - CEO - MikeNashTech.com

1 年

Wow, super newsletter and video explainer, Louis. AI is truly raising the bar of digital creativity. Great work. M

要查看或添加评论,请登录

Louis-Fran?ois Bouchard的更多文章

  • When is AI Integration REALLY Worth the Hype?

    When is AI Integration REALLY Worth the Hype?

    Good morning everyone! In this iteration, we discuss the biggest problem with AI: usefulness vs. pure marketing (hype).

    2 条评论
  • Advanced RAG Evaluation Techniques for Optimal LLM Performance

    Advanced RAG Evaluation Techniques for Optimal LLM Performance

    Good morning everyone! You are probably implementing RAG systems, missing out on easy improvements. Most don’t even…

    2 条评论
  • Indexing Methods for Vector Retrieval

    Indexing Methods for Vector Retrieval

    Good morning everyone! ChatGPT is pretty useful, but you can’t build a product out of it. I have been talking about RAG…

    2 条评论
  • Releasing our 90+ lesson practical LLM Developer course!

    Releasing our 90+ lesson practical LLM Developer course!

    I am super excited to finally announce that we ( Towards AI ) released our first independent industry-focussed course:…

    1 条评论
  • LLM Evaluations: Find the Best AI Model for Your Specific Task (no code)

    LLM Evaluations: Find the Best AI Model for Your Specific Task (no code)

    Good morning, everyone! Today, I’m excited to dive into an essential topic: why not all Large Language Models (LLMs)…

    4 条评论
  • Master Multi-Agent Systems Like a PRO with AGENTIC AI

    Master Multi-Agent Systems Like a PRO with AGENTIC AI

    Good morning everyone! In this iteration, I'm partnering with the amazing team at Integrail to make a really cool…

  • Running YOLOv7 on Your Phone

    Running YOLOv7 on Your Phone

    Good morning everyone! In this iteration, we are quickly reviewing how much the field of real-time object detection has…

    1 条评论
  • Building LLMs for Production now available everywhere!

    Building LLMs for Production now available everywhere!

    Good morning everyone! Today, I’m super excited to announce that "Building LLMs for Production" is now also available…

    2 条评论
  • A big Update for Building LLMs for Production!

    A big Update for Building LLMs for Production!

    Good morning everyone! Today, I’m super excited to announce that a new and improved version of Building LLMs for…

    18 条评论
  • Teaching AI to "Think"

    Teaching AI to "Think"

    Good morning, everyone! Like everyone else, we already talked about OpenAI's newest o1 model series, exploring how it…

    2 条评论

社区洞察

其他会员也浏览了