Field Reports

Leawood, Kan.

‍

‍Today, Torch.AI announced enhancements to their AI/ML-powered Nexus™ Unifying Synaptic Mesh (USM) platform that will deliver custom entity recognition models to production environments with TorchVision. As a field of AI that enables machines to derive meaningful information from digital images and videos, computer vision needs a lot of data to learn to “see.” To date, much has been invested in model development, however, with today’s announcement, Torch is solving the challenge of model deployment. Leveraging their highly scalable micro service technology, Torch Vision allows trained models to be quickly deployed to identify objects, people, text, scenes, and activities in image and video data. Nexus™ can now harness the flood of data being created from smartphones, security systems, traffic cameras and satellites, and provide structured and connected information across all datasets to customers.

‍

Real-world applications of computer vision are prominent insecurity-related fields, where mission speed is critical at massive scale. These include aiding warfighter recovery in hostile zones, locating victims in natural disasters, and detecting disease in agriculture or in the human body. Torch’s implementation of computer vision protects sensitive data through encryption and innovative deployment, where targets sought and tracked in video files are only revealed in weighted numbers on the models. The technology platform will process customers’ data streams, tag objects identified, and deliver results to customer analytics environments without users ever seeing the imagery—data protection and security at every step.

‍

Torch Vision provides an automated mechanism for object detection, action recognition, and target tracking within streaming video. These entities are tagged for enhanced search ability. These file types may contain hundreds of metadata properties, including geospatial, time-series, and environmental, making it possible to create connections across publicly available and third party data sources. By combining Torch Vision with existing data integration capabilities, Torch customers engage with one platform for decision-making.

‍