Attribute visualisation & Detection confidence
We previously released per polygon attributes for your results, we now extend this by adding a new attribute and attribute visualization tools to the platform.
Read articleSara Garbini
You are here:
Per accuracy area scoring
Per-accuracy area scoring is a feature that enhances the way users can assess the performance of their geospatial data detectors.
Read article
Confidence maps per accuracy area
The implementation of confidence maps is now available for accuracy areas. This feature provides users with an overlay on images that highlights areas where the model is uncertain about its detections, effectively visualizing where the model may be underperforming.
Read article
Training Warnings
This update extends and enhances the previously released localized warnings. This new functionality is designed to provide users with specific guidance on areas within their detector where the annotated ground truth and the model’s predictions do not align. It serves as a critical tool for identifying discrepancies and improving the overall accuracy of geospatial models.
Read article
Refined detection areas management
Navigate through existing detection areas, simplifying the inspection of results at scale and making direct edits as needed. Introducing the ability to create new detection areas directly from vector layers unlocking advanced workflows like chaining detectors together. Simplifing the drawing and editing process of new detection areas.
Read article
Band settings and split screen for results visualization
Effortlessly visualize different band setups within an image. Toggle between bands more efficiently, and compare two band settings side-by-side, such as an RGB and DSM image combining optical and height information images tailored for change detection.
Read article
Layers attributes
Now, the results from Picterra’s advanced GIS tools (such as “DSM height extraction”) can be visualized directly within the Picterra platform, bringing more comprehensive and detailed insights to the forefront. Boost your analysis by creating and editing attributes for detections within vector layers.
Read article
Mass selection
Introducing mouse-based mass selection, Supports mass delete functions and facilitates detection editing.
Read article
Magic wand
We are excited to announce the integration of Meta AI’s Segment Anything Model (SAM) into Picterra making the production of ML models faster, easier, and more accurate than ever before.
Read article
ArcGIS Image Server integration
We are excited to announce the expansion of Picterra’s range of supported imagery servers to include Esri’s ArcGIS Image Server! Available via ArcGIS Online, this update aims to provide a more seamless and efficient experience for our users, particularly those with large amounts of data stored on dedicated imagery servers.
Read article