Computer Vision Now: Full-Stack Visual for Advanced Image & Video Analytics

Chooch is an artificial intelligence-driven image and video analytics platform that provides accurate and complete computer vision. Chooch's advanced image and video analytics software delivers the entire AI lifecycle through a user-friendly dashboard - no coding required. With less than an hour of training, anyone can create and deploy accurate analytic models to the operational and computational edge.

Chooch is modeled on the human process of understanding the visual world in context, with deep learning neural networks that clone subject matter expertise directly into the intelligence production process.

Key Benefits

The Chooch platform reduces intelligence processing times from minutes to seconds. With rapid analysis and reporting, the initiative is returned to the analyst by automatically determining what an object is, which provides critical time to determine what it means. Chooch reduces human error and operator fatigue, and enables more accurate decisions in real-time.

Chooch eliminates third-party dependencies for dataset generation, annotation and labeling, and model training and testing, providing users complete control of their data. Once deployed, Chooch models automate object, facial, and text detection and recognition, geolocation, timing, and reporting to provide critical information to decision makers and enable immediate machine-to-machine tasking.

The rapid exploitation of visual data with Chooch’s system provides the persistent awareness needed to maintain an information advantage in emerging situations. Chooch can process multiple sensor feeds simultaneously and identify even subtle changes and patterns that can have significant impacts on intelligence assessments.

Deployment Advantage

Chooch leads the computer vision industry by reducing artificial intelligence model training and deployment times from months to minutes with precise detection and recognition throughout thousands of image classes in electro-optical, infrared, and synthetic aperture radar imagery and video. Chooch is hardware agnostic, but is optimized for NVIDIA graphics processing units, from Jetson to DGX systems.

Chooch is fluent in Electro-Optical, Infrared, and Synthetic Aperture Radar imagery and video (Static and Feed GEOINT). It encompasses an integrated network of neural networks with over 2,400 detection, recognition, and evaluation models, including change and anomaly detection functions. Chooch also provides over 250,000 public pre-trained object, image, and facial classes to minimize deployment time.

Security Advantage

Chooch's processing capacity can be scaled from a single edge device to unlimited terminals in the cloud or on-premises when security requires an air-gapped environment to secure critical data.

Chooch's federated edge network enables edge processors with Chooch software to connect to the customer's master cloud-based account, so users can push analytic models to their edge devices remotely, or pull the models from the cloud to the device. Either way, inferencing/processing occurs locally, and only critical data is communicated to decision makers when and how the user decides.

Chooch’s federated edge networks can be used to process satellite imagery on the satellite, within a local 5G network, through a video management system for facilities security, or sharing analytics across a multi-domain networks of sensors from sea, air, land, and space.