As the summer days swiftly pass by, it heralds the commencement of yet another exciting event – the OpenCV AI Competition! Every year, OpenCV collaborates with industry leaders to present a challenge to machine learning engineers worldwide, encouraging them to showcase their cutting-edge AI models and algorithms. What sets this competition apart is its open-ended nature; participants are free to explore any task or prompt, as long as the solution harnesses the power of the OpenCV Library, easily one of the best computer vision toolkits available today. From robotics to agriculture, education to health and medical, and even sports, submissions are diverse and innovative. The competition culminates on November 30th, granting ample time for participants to conceive extraordinary projects such as a laser-guided weed killer or an autonomous driving RC helicopter – the sky’s the limit!
Over the years, OpenCV has played a pivotal role in inspiring teams to step up and innovate to address real problems within their communities. The spirit of competition has a unique way of uniting people for a greater cause, driving them to exceed their ordinary limits. Events like this foster an atmosphere of growth, propelling the field of computer vision forward and sparking brilliant new ideas. As the deadline approaches, submissions will pour in from across the globe, showcasing the potential for groundbreaking breakthroughs. Some winners may spin off into new startups or, in the case of team DeViTech, who built a laser-based optical 3D scanner, will finally be able to balance their bookcase.
One such winner of the OpenCV competition, Cortic Technology, took this problem first hand by developing an AI platform for kids to incorporate computer vision with Legos! In this post, I’ll share what I learned from talking to Ye Lu from Cortic, and how open source FiftyOne could help you streamline your OpenCV project!
Cortic Technology’s Winning OpenCV Project
Cortic Technology, the triumphant team of the 2021 OpenCV Spatial AI competition, created a block programming language through a user-friendly interface, empowering young students to build their first computer vision applications. Their visionary founder and team lead, Ye Lu, aimed to minimize barriers for beginners in computer vision, enabling them to create incredible applications without prior expertise. When asked he explained that kids are hands-on learners and need to be immersed in the world of CV where they can see up close the results of their work. Cortic leverages things like Legos, Raspberry Pis, as well as mobile robots to construct a world where kids can be enchanted by modern AI developments. Also mentioned the need for powerful visualization tools, being able to see the detections, classifications, or other results in an interactive space can be extremely helpful for teaching kids.
The OpenCV AI Competition stands as a beacon of innovation, rallying talented minds from across the globe to push the boundaries of AI. It serves as a testament to the power of community-driven initiatives in fostering education and inspiring the next generation of AI enthusiasts. With the continued efforts of organizations like OpenCV and the remarkable endeavors of past winners like Cortic Technology, we are taking significant strides towards creating a world where AI literacy and responsible usage are commonplace.
How Open Source FiftyOne Can Help Enable Your OpenCV Submission
Want to turbocharge your OpenCV submission? Use open source FiftyOne to get the most out of your data! With the ability to store group data, different modalities, and present in intuitive and insightful ways, FiftyOne is a force multiplier to your project. Let’s go back to the autonomous driving RC helicopter idea from the opening paragraph. What if we were going to hover around our house to track migrating butterflies? Luckily with FiftyOne, data has never been easier to curate and visualize! We can pop a butterfly dataset into FiftyOne and voila!
FiftyOne can empower you to extract stunning insights from your dataset. You can add detections, run evaluations, compute embeddings, and much much more! Once your data has been uploaded to FiftyOne, a world of options open up for you. You can do uniqueness sorting to find your most unique butterflies in your dataset like below.
If you are a beginner to not just FiftyOne but all of computer vision, you can do simple operations as well such as finding your favorite butterfly type. To find more instructions on how you can gain control over your data, please head over to all of our tutorials here!
Want help on how to unlock FiftyOne’s potential for your OpenCV AI solution? Join the FiftyOne Community Slack and find others taking initiative to tackle some of the world’s most interesting Computer Vision problems. Best of luck to all entries to the competition!
Join the FiftyOne Community!
Join the thousands of engineers and data scientists already using FiftyOne to solve some of the most challenging problems in computer vision today!
- 2,000+ FiftyOne Slack members
- 4,000+ stars on GitHub
- 5,000+ Meetup members
- Used by 370+ repositories
- 60+ contributors