Open Source Reference Implementations
Deploy your own IoT solutions by using these prebuilt open source projects.
Use computer vision solutions to adjust kiosk displays in response to audience demographics.
- Determine age, gender, and head pose with deep neural network (DNN) models.
- Play a 4K ad based on audience identification.
Deploy sensor fusion technology for an automated checkout that enables real-time insight about the products consumers are buying using the EdgeX Foundry* extensible framework.
- Authenticate and authorize different users
- Develop add-on services and sensors
- Recognize items, detect discrepancies, and record real-time data
Detect loss at self-checkout by seamlessly connecting different sensor devices, including weight scale sensors, cameras, and RFIDs.
- Recognize products entering and exiting retail checkout areas
- Develop add-on services and sensors
- Use multiple edge sensors to accurately recognize items, detect discrepancies, and record a real-time transaction log (RTTL)
Build a solution to analyze customer expressions and reactions to product advertising collateral that is positioned on retail shelves.
- Measure active versus inactive user product engagement.
- Capture analytics on shopper reactions to visual ads.
Detect the mood of shoppers as they look at a retail or kiosk display.
- Mall shoppers using interactive or map kiosk
- Grocery store shoppers viewing digital signage ads
- Hospitals using a kiosk to assist patients or visitors
Use computer vision inference in the Intel® Distribution of OpenVINO™ toolkit to provide analytics on customer engagement, store traffic, and shelf inventory.
- Detect people within a designated area by displaying a green bounding box over them, count the total number of people, and the time they are in the frame.
- Use an inferencing pipeline to detect faces, emotions, and head poses.
Capture video, generate a heat map, record the number of people present, and then integrate the results. The program can also create an output video and save snapshots.
- Factory or warehouse activity
- Restaurant activity
- Hospital lobbies and floors
Monitor three different streams of video that count people inside and outside of a facility. This application also counts product inventory.
- Movement of people
- Foot activity in retail or warehouse spaces
- Inventory availability of products on shelves
Segment brain tumors in raw MRI images by applying the U-Net architecture.
- Detect brain tumors in MRI images.
- Plot predictions from segmented brain tumors.
- Predict results using a pretrained model and the Sørensen–Dice coefficient.
Detect pneumonia in X-rays using computer vision inferencing and a pretrained model.
- Predict the probability of infection caused by pneumonia.
- Identify anomalies and predict results with medical imaging.
- Train models for classification using labeled X-rays from open source datasets.
Run multiple independent anomaly detection workloads on a single system that runs multiple virtual machines through a Kernel-based Virtual Machine (KVM) host.
- Industrial computer vision and time-series analysis
- Virtualized environments for multiple applications
Predict performance issues with manufacturing equipment motors. Perform local or cloud analytics of the issues found, and then display the data on a user interface to determine when failures might arise.
- Air conditioning units
Detect various anomalies of an object that is moving on a conveyor belt within a manufacturing facility, and then run analysis on what is detected.
- Identify products damaged during manufacturing.
- Confirm product orientation is correct for labeling.
- Detect labels for product UPC identification.
Use computer vision to detect and measure the approximate size of parts moving on an assembly line, identify irregularly sized items, and send alerts when any are detected.
- Track mechanical part count and size.
- Alert users if an irregularly sized part is detected.
- Interpret data from either a live webcam or preexisting video.
Create a concurrent video analysis pipeline featuring multistream face and human pose detection, vehicle attribute detection, and the ability to encode multiple videos to local storage in a single stream.
- Retail digital surveillance such as network video recorders
- Video matrix commercial multimedia applications
- Video conference multipoint control units (MCU) and terminals
Implement and use Intel® hardware platforms for video decoding, encoding, and optimization using various media stacks.
- Transmit video into a computer vision application for people detection.
- Use GStreamer and the Intel® Media SDK to capture video streams and encode them into a format that can be stored on a server.
Build an application that alerts you when someone enters a restricted area. Learn how to use models for multiclass object detection.
- Record and send alerts on activity in controlled spaces.
- Track parking lots, entrances, and property.
Receive or post information on available parking spaces by tracking how many vehicles enter and exit a parking lot.
- Track and analyze vehicle activity.
- Report on parking space availability.
Create a smart video application using the Intel Distribution of OpenVINO toolkit. The toolkit uses models and inference to run single-class object detection.
- Track activity in retail.
- Observe factory work spaces and building entrances for activity.
- Capture and record information on the number of people.
Secure work areas and send alerts if someone enters the restricted space.
- Track worker activity in proximity to heavy machinery.
- Develop safety solutions using computer vision technologies.
Observe workers as they pass in front of a camera, identify them using facial recognition, and determine if they have adequate safety protection.
- Ensure safety in the industrial workplace.
- Detect the presence of required safety equipment.
- Monitor factories, warehouses, etc.