Pick a Pic is an open-sourced application for collecting human feedback on AI-generated images for academic research in AI. The app allows users to generate images from text and then rank them.

All code, data, and models resulting from the research project are open-sourced and available to the public. Users must authenticate via Google before generating and participating in the project.

The mission is to build the largest open-sourced and publicly available human-feedback for text-to-image dataset. Currently, the app uses Stable Diffusion 2.1 and ProtogenX3.4 models.

The data will be updated periodically and the collected data will be released into the public domain. The project is supported by the Google TRC Program.

Visit website

Save

Featured

Feature requests

Are you looking for a specific feature that's not present in Pickapic?

💡 Request a feature

Pickapic was manually vetted by our editorial team and was first featured on January 10th 2023.

Promote this AI Claim this AI

DeepFiction AI

Fictional stories

Unleash Creativity with Uncensored Stories

★★★★★

★★★★★
(8)322
7

From $10/mo
Share

B12.io

Website building

AI-powered website builder

★★★★★

★★★★★
(10)637
3

From $42/mo
Share

Venturekit

Business plans

Business plan generation and consulting services

★★★★★

★★★★★
(17)800
10

Free + from $16/mo
Share

453 alternatives to Pickapic for Image generation

Most impacted jobs

Pros and Cons

Pros

Open-sourced application

Image generation from text

Collection of human feedback

Image ranking feature

Publicly available output

Google authentication

Largest text-to-image dataset

Stable Diffusion 2.1 model

ProtogenX3.4 model

Periodically updated data

Released into public domain

Google TRC Program support

Access to code repository

Ability to participate

For research purposes

Potential for academic collaboration

Dedicated community support

Clear project mission

Structured model inference repository

Chat support via Discord

Cons

Requires Google authentication

Data collection consent required

Only uses Stable Diffusion 2.1 and ProtogenX3.4

Periodic update of dataset

Publicly releases user data

Project support dependent on Google TRC Program

Q&A

What is the main purpose of Pick a Pic?

The main purpose of Pick a Pic is to collect human feedback on AI-generated images for academic research in AI.

How does Pick a Pic generate images?

Pick a Pic generates images using AI models. Specifically, it uses Stable diffusion 2.1 and ProtogenX3.4 for the generation process.

How can users contribute to the Pick a Pic project?

Users can contribute to the Pick a Pic project by generating images from text, and then ranking them in order to provide valuable feedback.

What academic research is Pick a Pic used for?

Pick a Pic is used for academic research in the field of AI, specifically focusing on AI-generated images.

Which models does Pick a Pic use to generate images?

Pick a Pic uses Stable diffusion 2.1 and ProtogenX3.4 models to generate images.

How often is the data on Pick a Pic updated?

The data on Pick a Pic will be updated periodically, although a specific timeframe is not given.

Is all the code and data from Pick a Pic open source?

Yes, all the code and data from Pick a Pic is completely open-sourced. This includes both the app's repository and the model inference repository.

How can users authenticate to use Pick a Pic?

Users are required to authenticate via Google before generating and participating in the Pick a Pic project.

What information is collected from users in Pick a Pic?

Pick a Pic collects the prompts, interactions, and generated images from users for research purposes.

Are there any restrictions on how the Pick a Pic demo can be used?

The Pick a Pic demo is intended for research purposes only, and the authors disclaim responsibility for any misuse of the demo.

How can users provide feedback on the images generated in Pick a Pic?

Users can provide feedback on the images generated in Pick a Pic by ranking them after they are generated.

How can the collected data from Pick a Pic be accessed?

The collected data from Pick a Pic can be accessed from the project's dataset page on huggingface.co.

Can anyone use the images generated by Pick a Pic for their own purposes?

The images generated by Pick a Pic will be collected and eventually released into the public domain. This suggests that anyone can use them for their own purposes.

Can I contact the Pick a Pic team?

Yes, the Pick a Pic team can be contacted through discord or by email.

What does 'Negative Prompts' mean?

"Negative Prompts" refers to the prompts enclosed within square brackets. For example, "Hello [world]" will provide the main prompt 'Hello' and the negative prompt 'world'.

How is the project, Pick a Pic, supported?

The project, Pick a Pic, is generously supported by the Google TRC Program.

Where can I find the code and models used by Pick a Pic?

The code and models used by Pick a Pic can be found on Github and Hugging Face, respectively.

Why am I required to authenticate via Google for Pick a Pic?

Authentication via Google is required for Pick a Pic to ensure that real humans are participating in the information collection.

Do I need special permissions or clearances to use Pick a Pic's data or code?

No, there's no mention about needing special permissions or clearances to use Pick a Pic's data or code. However, proper referencing and citation would be expected by most open-source communities.

What happens if something goes wrong while using Pick a Pic or there are too many users?

If something goes wrong while using Pick a Pic or there are too many users, it recommends refreshing the page and trying again.