A team of PhD student researchers from Saudi Arabia has developed a new AI-powered tool, MiniGPT-4, with attributes similar to OpenAI’s ChatGPT-4.
Since ChatGPT’s November release and worldwide hit, developers have been doing everything they can to come up with new AI tools that can rival or complement popular chatbots.
MiniGPT-4, developed using the ChatGPT model, is just the latest example.
Also read: Bill Gates: AI chatbots can teach kids to read and write in 18 months
according to future toolsMiniGPT-4 can perform many tasks such as generating image descriptions and building websites.
“The tool generates detailed image descriptions, creates websites from handwritten drafts, writes stories and poems inspired by given images, and provides solutions to problems illustrated by images. serve and teach users how to cook based on food photos,” claims Future. tool.
When ChatGPT-4 was released, it showed a video of a model building a website from sketch images. Bercy, MiniGPT-4 is capable of accomplishing the same feat. The only difference is that ChatGPT-4 is not available to everyone at the moment and MiniGPT-4 is already popular.
About MiniGPT
according to Gax, MiniGPT-4 uses an advanced LLM called Vicuna as a language decoder. It builds on LLaMa and is reported to achieve 90% of ChatGPT’s quality as rated by GPT-4.
The AI model uses Bootstrapping Language Image Pre-training (BLIP-2) pre-trained components and adds a single injection layer to freeze all other visual and linguistic components into encoded Aligned the visual features with the Vicuna language model.
David Watson He said MiniGPT is lightweight and easy to implement in real-time situations such as chatbots, virtual assistants, and automated image captioning systems.
He also cites some applications where MiniGPT-4 can be used successfully. Image descriptions for the visually impaired using audio descriptions, how text-to-speech systems should be included.
in the meantime Open AI We have confirmed the multimodal capabilities of GPT-4, but have not yet released the image processing capabilities. MiniGPT-4 fills this gap by using a more sophisticated LLM to process images along with language.
AI tools to support research
Experts say the state-of-the-art underlying language model used is designed to help researchers advance their research in this particular AI segment.
Given that OpenAI has not disclosed much information about GPT-4’s architecture, model size, hardware, training compute, dataset construction, or training methods, the open-source nature of MiniGPT-4 It may prove particularly valuable to researchers.
“MiniGPT’s ability to process images offers researchers new opportunities to investigate the relationship between language and visual models,” writes Yana Khara. Analytics Vidaya.
“MiniGPT-4 can drive innovation and advancement in AI technology by providing a smaller, more accessible model for researchers.
“Furthermore, the model’s open-source foundation allows the research community to collaborate and share their findings to make further progress in this area.”
MiniGPT takes image captioning to another level
Baseewho tweeted a thread explaining how to use the MiniGPT-4 to chat with images included some of the following cases:
Fix broken items
After uploading a picture of the broken item to the MiniGPT platform and asking how to fix the situation in the image, the chatbot will explain the situation in the image and suggest ways to fix the identified issues.
in the tweet, MiniGPT easily identifies washing machine leaking problems, explains why leaks occur, and provides a list of solutions that users can try.
write an ad
another Tweet from Bercy A MiniGPT thread involved a scenario where a photo of a mug that a user had created and sold was passed to MiniGPT. The user then asks the chatbot to write an ad to sell mugs. Chatbots do this well.
brief introduction
Just upload a picture of your movie and ask MiniGPT for a brief introduction. Next, generate a paragraph introduction for the movie in question.as seen in tweet, The MiniGPT chatbot recognizes the image of “The Godfather” and writes the movie intro according to your instructions.
Since the launch of ChatGPT, the market has developed a myriad of new AI tools. There are other alternatives to the famous chatbot, and others have reportedly surpassed them. Auto-GPT in particular is still making waves in the AI community. As things stand, it seems almost inevitable that we will be baffled by the wealth of AI in virtually every human task.
