Available for Advanced, Teams, and Enterprise users during open beta
Note: After the beta period, AI Annotations will become an add-on feature.
Read more about AI Annotations here!
Read more about On-Demand Custom Models for Enterprise Customers here!
Custom Models for AI Annotations:
The custom objects in AI Annotations feature is designed to save users valuable time when annotating large volumes of data. It leverages custom object classes of AI Annotations to automatically detect and annotate specific elements on your maps, streamlining your workflow and improving efficiency.
Currently, there are three types of output for the custom models: Area, Volume, and Count annotations.
The Custom Models feature is designed to save users valuable time when annotating large volumes of data. It leverages advanced AI models to automatically detect and annotate specific elements on your maps, streamlining your workflow and improving efficiency.
Available Custom Models & Intended Use Cases:
Currently, the following AI models are available for use:
Counting: | BEST Use Case: | NOT Intended For: |
Counting Orchard Trees |
Discrete, single-standing orchard trees preferably in the nursery environment : |
Individual non-orchard trees (shrubs, pine trees, etc.). Canopy trees, overlapping forest trees (best used with the Find Forest Trees model):
|
Counting Palm Trees |
Actual palm trees: |
Other types of trees standing separately (bush, shrub, etc.): |
Counting Solar PV Modules |
Utility solar panels installed in solar farms with a good level of image overlap and high-quality processing: |
Solar panels installed on commercial and residential roofs: |
Counting Cars |
Recognizing passenger cars (sedans, suv's, and pickup trucks) in parking lots within an urban environment:
|
Recognizing industrial vehicles and non-passenger cars on jobsite environments:
|
Finding Stockpiles |
Distinct single material stockpiles and clearly distinguishable: |
Mixed Material stockpiles, trash, and removed piles: |
Finding Forest Trees |
Mixed trees, overlapping canopies, and green and live trees: |
Timber, bush, and fallen trees: |
We are constantly working on expanding our selection of Custom Models, and new models will be made available as soon as they are ready.
Area Annotations
If the AI model is an Area model, the output will consist of area annotations. These annotations will calculate the total amount of distinct areas where AI detection was found to be true. This is particularly useful when you need to identify and measure various objects or features on your map.
Count Annotations
For Count annotations, the AI model will produce a single annotation indicating the total number of detected objects. This type of annotation is helpful when you need to quickly count the occurrences of specific objects on your map.
Frequently Asked Questions
How do I train a custom model?
To train a custom model for your Organization, please fill in the custom model request form. Our team will review your request and do their best to create a custom model that suits your needs. Please make sure, you indicate all the fields in the form, specifically the ones with the examples. We will not be able to respond to your request unless we have examples of objects on a map.
Can I run multiple models at the same time?
Yes, you can run multiple models simultaneously. We do not restrict users from running multiple models simultaneously to maximize efficiency.
Can I run multiple models on the same map?
Absolutely! You can run as many models as you want on the same map. However, please be cautious not to duplicate model runs to avoid unnecessary annotation duplication.
How long is the wait time for the Custom Models?
The AI Annotations Tool is designed to deliver results rapidly. In most cases, annotations are completed in under 5 minutes. However, for exceptionally large maps, the processing time may extend up to 20 minutes. Rest assured, once the job is complete, you will receive an email notification with the results.