The challengeCreate a proof of concept using machine vision on embedded systems to address an industry-grade problem.
For the first time, the tinyML Foundation has launched a contest for developers worldwide, challenging them to build advanced applications with low-power machine learning inferencing and computer vision for edge devices.
Tiny machine learning (tinyML) is broadly defined as a fast-growing field of machine learning technologies including hardware, algorithms and software capable of performing on-device sensor data analytics at extremely low power, typically in the mW range and below, and hence enabling a variety of always-on use cases and targeting battery-operated devices.
Here's where you come in.
We're calling on developers everywhere to devise a proof of concept using machine vision on embedded systems to address an industry-grade problem. Submissions can leverage any embedded vision hardware platform and software framework. Extra points for battery-operated, ultra-low-power projects.
Models deployed on hardware like the OpenMV Cam H7, Arduino Portenta Vision Shield, Luxonis LUX-ESP32, Himax WE-I Plus, Arducam Pico4ML, and Raspberry Pi are ushering in a whole new era of solutions that can now accurately identify and classify what they see, across a multitude of workloads:
Select from any of these suggested categories or choose your own area of interest. Submissions will be awarded points based on their ability to address a problem within a given industry.tinyML Summit 2021 announcement
Watch Edge Impulse co-founder Zach Shelby and OpenMV co-founder Kwabena Agyeman introduce the challenge and offer some sample applications to help get you started.Executive sponsorsPlatinum sponsorsPlatform partners
We are giving away $6,000 in prizes to the top three submissions. Judges will score your projects based on the judging criteria outlined in the rules tab.
Hackster Impact Prize
Join Hackster as we rally with the United Nations’ Sustainable Development Goals (SDGs) in order to make today's world a smarter, healthier place. The Hackster team will select five projects that best contribute to any of the SDGs. Please refer to the specific judging criteria for this prize category.
The winner will receive the following:
- a $250 gift card;
- a 12-month subscription to Patchr Premium ($180 value);
- last but not least, they will be featured in a video interview with Hackster’s one and only Alex Glow.
Please visit the Hackster Impact Prize page for more information.
Co-founder of OpenMV, a company dedicated to making machine vision on microcontrollers easy. The OpenMV Cam makes it possible for students and hobbyists to build simple robots that can track colored objects or faces. For example, with the OpenMV Cam and a Nerf Gun engineering, major college students can build a color tracking turret for a mechatronics capstone project.Jay Burris
Academic Program Manager at Intel who has been working in the embedded, automotive, and Internet of Things industry for 20+ years. He is passionate about advocating for AI developers, students and faculty to help them advance their knowledge using Intel Technologies such as the AI toolkit called Intel®️ Distribution of OpenVINO™️ toolkit, learning through Intel’s new Edge AI Certification program, and prototyping ideas inside the cloud developer environment Intel® DevCloud for the Edge.Evgeni Gousev
Senior Director, at Qualcomm responsible for leading HW/SYS research and development, ultra-low power edge computing platform and machine vision, tinyML foundation member and advocateSamuel Kuo
Senior Director, R&D for ASIC development at Himax Technologies Inc., with responsibility for ultra-low power Endpoint AI processor and ecosystem development.Zach Shelby
An entrepreneur, angel investor and technologist in the embedded space with a passion for TinyML, embedded and Internet engineering. Zach is a former Arm VP, founder and CEO of the Micro:bit Foundation and Sensinode, active in several of his portfolio companies, and is working to bring ML to any embedded device as co-founder and CEO of Edge Impulse.
Complimentary & Discounted Hardware
In order to accelerate your development, tinyML Vision Challenge sponsors have made complimentary and discounted hardware available. Here is what is currently available:
Intel and Luxonis are pleased to offer a 50% discount on a DepthAI LUX-ESP32 to the first 50 project idea submissions. We want to hear your ideas! $100 discount to LUX-ESP32 DepthAI Embedded SOM with WiFi, BT, IMU, and On-Board Cameras ($99 after 50% discount). Provide a description of a real-world spatial AI computer vision problem and proposed solution. Use of AI with the OpenVINO™ toolkit and spatial depth sensing are encouraged. Please submit your project idea proposal, detailing how your solution idea will use the Luxonis LUX-ESP32 DepthAI Embedded Camera using OpenCV and AI acceleration with the Intel® Distribution of OpenVINO™ toolkit. Submissions are due by July 8th, 2021 at noon PDT. Must be registered for the tinyML Vision Challenge to qualify for discount, and be one of the first 50 to submit a project idea. Submit your request for a discount while supplies last!
Pixart Imaging has a limited supply of complimentary PAJ6100U6 OpenMV Shields for use on the Open H7 platform. To request a shield please send email to email@example.com with the subject "PAJ6100U6 OpenMV Shield".
- Autonomous Driving with OpenMV and Edge Impulse
- Arm Innovation Coffee: tinyML on OpenMV with Edge Impulse
- Arm Innovation Coffee: Embedded Machine Vision with Arduino Portenta
- Supporting Tensorflow Lite MCU in tiny low power FPGAs
- Intel & Luxonis AI-Powered Backpack Helps Visually Impaired Navigate World (also article)
- Luxonis YouTube channel
- OpenVINO™ toolkit video series and YouTube channel
Tutorials & Guides
- Connecting the OpenMV Cam H7 to Edge Impulse
- Connecting the Himax WE-I Plus to Edge Impulse
- Getting Started with the Himax WE-I Plus
- Connecting the Portenta and Vision Shield to Edge Impulse
- Connecting the Eta Compute ECM3532 AI Vision Board to Edge Impulse
- Getting Started with Lattice sensAI
- Object detection on Lattice FPGA
- Predictive maintenance on Lattice FPGA
- Learning more about DepthAI
- Getting Started with DepthAI
- Luxonis Open Source Hardware Designs on Github
- Learn and Develop AI with the OpenVINO™ toolkit using the Intel® DevCloud for the Edge… free registration and access
- Download the OpenVINO™ toolkit
- AI Vision and Sensor Fusion with Himax WE-I Plus
- Ultra-Low-Power Computer Vision Applications on the Portenta
- Ultra-Low-Power Computer Vision Applications with Edge Impulse and the Eta Compute AI Vision Board
The inaugural tinyML Summit in March 2019 showed very strong interest from the community with active participation of senior experts from 90 companies. It revealed that: (i) tiny machine learning capable hardware is becoming “good enough” for many commercial applications and new architectures (e.g. in-memory compute) are on the horizon; (ii) significant progress on algorithms, networks and models down to 100kB and below; and (iii) initial low power applications in the vision and audio space. There is growing momentum demonstrated by technical progress and ecosystem development.
Gain reputation points!
Earn reputation points for participating in this contest:
- Submit a project entry: 10 points
- Submit a project entry 30 days before the deadline: 50 points
- Be among the first 10 to submit a project entry: 50 points
- Your submitted project entry wins the contest: 250 points
April 15, 2021 at 8:00AM PT
September 17, 2021 at 11:59PM PT
Winners announced by
Oct 1, 2021