![Scalio](/img/logo-animated.png)
LexSet is a visual AI solution for object recognition and visual search. LexSet's Synthetic Data utilizes their rendering engine built from 3D visualization and simulation technologies to generate robust training sets that can complement existing training sets or serve as the foundation for a customer’s machine learning models. They take 3D CAD files and 3D scans of objects to generate thousands of images of each object, in all possible lighting states, occlusions, & camera angles for an ideal training set, for almost any object recognition application. LexSet has already generated over half a billion synthetic images. Their custom synthetic training datasets allow real-time attribution and SKU-level object recognition.
LexSet’s mobile cross-platform SDK is a RESTful API that can be utilized by mobile applications on any platform. The SDK enables real-time recognition, 3D tracking, and contents listing. The tool can also be used to help create digital twins for virtual room planning / virtual scene creation.
Scalio has partnered with and invested in LexSet to originally help pioneer spatial-semantic search using deep convolutional neural networks for multi object detection in AR, under patents created by Bill Gates and Satya Nadella (Microsoft).
Patents are as follows:
- US 8,708,223
Systems and methods for scanning a user environment and evaluating data of interest - US 8,959,082
Context-sensitive query enrichment - US 9,141,188
Presenting an augmented view in response to acquisition of data inferring user activity - US 8,341,522
Enhanced contextual user assistance
Objectives
Technologies
Accolades
WHAT THEY SAY
We partnered with Scalio because of the track record of their team. Due to the bleeding-edge nature of the technology stack, we knew Scalio was the right team to turn to for helping us push our ideas into reality.
- Les Karpas, President & Co-Founder at Lexset