Automated Machine Learning Pipelines for Object Recognition
Train custom machine learning models, and use them to change the way you think about your image data, all within MainEvent.
A picture’s worth a thousand words, but the right few words make all the difference.
With MainEvent Custom Models, you can train object recognition models on the brands and products you’re already tracking in MainEvent.
- Bucket “Bud Light Chelada + Tajin”
- Can “TAJIN CHILE LIMON” Bud Light Chelada
- Koozie “Bud Light Chelada + Tajin”
Go from using images to demonstrate proof of life, to extracting tangible and reportable data. It takes less than 1 second to process an image in realtime, or batch process hundreds of thousands of images over night, all behind the scenes.
Extract brand and product-specific insights.
↓ ↓ ↓
Detected Bucket “Bud Light Chelada + Tajin” with confidence 85.67% at location [332.40, 1682.35, 1091.66, 2453.75]
Detected Can “TAJIN CHILE LIMON” Bud Light Chelada with confidence 79.75% at location [2118.77, 1630.58, 2438.07, 2241.23]
Detected Can “TAJIN CHILE LIMON” Bud Light Chelada with confidence 79.64% at location [1178.51, 1641.47, 1437.07, 2306.95]
Detected Can “TAJIN CHILE LIMON” Bud Light Chelada with confidence 78.82% at location [1799.10, 1634.25, 2079.00, 2260.98]
Detected Koozie “Bud Light Chelada + Tajin” with confidence 78.66% at location [2032.73, 2232.12, 2594.17, 2689.67]
Detected Can “TAJIN CHILE LIMON” Bud Light Chelada with confidence 75.81% at location [1528.06, 1640.90, 1782.65, 2282.72]
Detected Can “TAJIN CHILE LIMON” Bud Light Chelada with confidence 70.39% at location [487.44, 1598.93, 774.69, 1980.64]
Detected Koozie “Bud Light Chelada + Tajin” with confidence 65.10% at location [1347.59, 2289.29, 1781.68, 2739.22]
Detected Koozie “Bud Light Chelada + Tajin” with confidence 64.29% at location [1687.79, 2260.78, 2185.96, 2722.33]
Detected Can “TAJIN CHILE LIMON” Bud Light Chelada with confidence 62.98% at location [814.48, 1578.37, 1072.46, 1943.83]
Describe the image in natural language.
Along with detection data, MainEvent generates a digestible narrative of the contents of the image.
The image prominently features a “Bucket Bud Light Chelada + Tajin” along with several cans labeled “TAJIN CHILE LIMON Bud Light Chelada”. Additionally, there are multiple “Koozie Bud Light Chelada + Tajin” items dispersed throughout the image.
Summary:
- 1 “Bucket Bud Light Chelada + Tajin”
- 6 cans of “TAJIN CHILE LIMON Bud Light Chelada”
- 3 “Koozie Bud Light Chelada + Tajin” items.
Build it into the rest of your MainEvent processes.
- Product Identification: Identify customer and competitor products accurately.
- Product Placement: Confirm placement instructions are being followed, and to what degree.
- Customer Identification: Happy and are holding the product correctly? Label for inclusion in summary and compliance reports.
- Inventory Tracking: Identify and count the items you care about.
- Automated Checklist Generation: Making the process of creating checklists easier to match the campaign requirements for setup and execution.
“We handle extensive field marketing campaigns with MainEvent. It's our source of knowledge, how we analyze and influence outcomes, manage time claims, and automate as much of our processes as possible…image recognition and everything ML is super exciting as we continue to grow and look forward.”
Alexa Reynoso, Sr. Account Director
How can machine learning and object recognition transform your field marketing data? Book a quick demo - we’d love to brainstorm with you.
