🚀 Introducing `hailo-apps`: GenAI at the Edge & A Massive Library Update!

We’re thrilled to announce a major evolution in how you build with Hailo. We’ve streamlined our repository structure to make developing and deploying advanced AI applications easier and more unified than ever before.

:package: The New Home for Applications: hailo-apps

We have officially renamed and upgraded our main application repository.
The new main entry point is now:
:backhand_index_pointing_right: https://github.com/hailo-ai/hailo-apps

This repository is your central hub for official Hailo applications. It consolidates our infrastructure and examples into one robust location, whether you are developing on x86 or embedded platforms like the Raspberry Pi.

:sparkles: The GenAI Era is Here

The biggest highlight of this release is the introduction of Generative AI support. We are pushing the boundaries of edge computing by enabling you to run powerful GenAI tasks directly on Hailo accelerators.

  • Vision Language Models (VLMs) & LLMs: You can now explore examples that integrate Large Language Models and visual understanding.

  • Voice Assistant: Check out our new fully functional voice assistant example, demonstrating conversational AI running at the edge.

  • Agent Tool Usage: Empower your agents to interact with the physical world! We’ve added examples where GenAI models initiate Python-based tools to control real hardware like servos and LEDs.

  • Interactive Applications: We have added new examples that demonstrate how to build responsive, intelligent interfaces using Hailo’s efficient inference.

:rocket: A Growing Library of Applications

Beyond GenAI, this release includes a significant expansion of our application catalog. We have added lots of new applications covering a wide variety of use cases.

  • Python & C++ Examples: We have expanded our language support to include a rich collection of both Python and C++ based examples, giving you the flexibility to choose the best environment for your production needs.

Whether you are looking for classic computer vision tasks to LLMs, the new hailo-apps repository is packed with ready-to-use examples to help you get to production faster.

:books: Read the Full Release Notes

For a complete list of all the new applications, features, and improvements, please check out the release notes:
:page_facing_up: Release Notes

:hammer_and_wrench: Give It a Try!

Head over to the Hailo-Apps, clone the code, and start exploring the new GenAI capabilities.

We can’t wait to see what you create! Start building, share your projects with the community, and let’s push the boundaries of innovation together!

Happy Coding!
The Hailo Team

4 Likes

In this comment @giladn said there would be an LPR pipeline released in the next release of hailo-apps, but I can’t find anything about that in the source code Could you show us where that ended up and how to use it?

I would use the existing examples published in the model-zoo but those are unusable due to a broken link that @omria said would be fixed in this comment from September. However The link is still broken and says ‘access denied’. I have a hailo developer zone account and it’s not available through there either.

1 Like

Hi @dreamily9869 , we’ve introduced several application and infrastructure changes in this release, and as a result, the LPR pipeline wasn’t ready in time for launch. We are actively working on it and expect it to be available within the next few days. In the meantime, if you don’t require real-time video processing, I highly recommend checking out this excellent tutorial by @shashi: A Comprehensive Guide to Building a License Plate Recognition (LPR) Systems

Looking for llama_3_2_1b_instruct_hailo8.hef file to use with my Pi 5. Please help me find it!