Company Performance Metrics
Gum launches niche cloud-based computer vision/AI products for unstructured scene analysis in images/videos.
Product: grepCV Multi-tenant, multi-purpose Serverless microservices architecture for processing high-resolution natural images/video (unstructured scenes), extracting metadata via Computer Vision, and indexing resulting
metadata associations for search. The system has both a receipt API (for initiating new workflows) & a request API (to search/retrieve processed objects by their associated metadata).
A Custom set of libraries work together to provide OCR for unstructured scenes (natural images) via a 3-pass recognition routine. grepCV accuracy is comparable on pages of text to commonly available libraries/APIs (Google Vision, Tesseract, OCRopus, etc.), but exceeds accuracy on unstructured scenes at comparable performance.
Specialties Computer Vision, Cloud Architecture, Machine Learning, OCR, Microservices
