I am excited to share ‘TinyML Paper and Projects’ repo with the community. It mainly consists of a few collections of research paper(from 2016 to date). Along with a couple of awesome TinyML projects and blog posts. We hope to add more content to it. If you have any suggestions or resources related to TinyML, please feel free to reach out on this thread.
Thank you for providing this most helpful resource. The various papers, presentations, and blog posts are extremely beneficial to me as I am a beginning novice in TinyML and machine learning in general. Many thanks for taking the time to assemble this collection and make it available to all.
Hi @gigwegbe !
I am Baptiste Zloch. Today I finished my 1st Tiny ML project, from data collection and processing to model deployement on nano33 board.
Know what ? it works ! I am so enthusiastic !!
Here the link to the whole project (model, .ipynb, c++…) of course you can use it.
@BaptisteZloch That’s awesome, would you up for sharing a video of your project under the following?
Nothing beats a demo
Here are some other projects:
Thanks for sharing your awesome project, I have added it to the repo.
I can’t wait to see your demo. Cheers
Thanks I am glad of your message ! I don’t know when, but I can share a demo of my project on a video of course !
New interesting paper: https://arxiv.org/pdf/2102.01255.pdf
Thanks for posting this awesome resource! It is great to have all these papers and projects in one place.
I have one suggestion–related to organization rather than TinyML itself. You already organized papers/projects into categories. As the list gets longer it will be beneficial to add more structure. Maybe sort papers by category or citation count? I picture a tabular format with Category (Ex. Benchmarking, Architectures, Hardware, Algorithms), Title, and Year.
GitHub Readme.md are markdown and don’t sort. So the lists would live in a spreadsheet and then paste into Readme.md. Or maybe sorting is done by the user in Google Sheets instead of the Readme.md. Hope this suggestion is useful.
Thanks for all the time spent to assemble this information!
New paper: https://arxiv.org/pdf/2103.08295.pdf