Introduction

OMNI Agent is a video analytics API module designed for detecting faces, silhouettes, human movements, and specific objects in camera streams or video files, with the capability to synchronize with OMNI Platform or third-party services via webhook and MQTT.

Key features

Detection and tracking of faces, bodies, human skeleton key points, and specific objects

Recognition of gender, age, emotions, and other facial attributes

Face identification (1:N) by control lists including masked or partially covered faces

Behavioral analytics: detection of falls, fights, and people lying or sitting

Detection of a human or object’s presence within a region of interest (ROI) or when a human or an object crosses a virtual line in the frame

Detection and tracking of arbitrary objects using a pair of frames through Few-Shot Learning technology