Object Detection — often referred to as object recognition — engines in the Veritone cognitive engine ecosystem detect the presence of one or multiple objects within video or still images, including vehicles, people, animals, and more. For example, satellite images can be automatically reviewed for anomalies, or hazards automatically detected on roads and runways.
Object Detection Features:
- Trainable with Custom Libraries
Create custom models using objects in images or video specific to your use case with the Veritone Library application or annotation tool.
- Broad Pre-Trained Object Library
Identify thousands of objects including vehicles, people, animals, landmarks, weapons, colors, food, art, apparel, and more in video and images with pre-trained labeling.
- Searchable Results
Identify the objects you are looking for quickly by label within images and video with searchable object detection engine output via API and Veritone applications.
- Near Real-Time Processing
Process pictures and video files in near real-time for use cases requiring fast turnaround of objects detected in data sets such as broadcasts, CCTV, still image collections, and more.
- Files or Stream Support
Recognize objects in short-form or long-form videos in video recordings, streamed video recordings or images, or live video or image streams.
- Flexible Deployment
Deploy in a new or integrate into an existing application in the cloud via aiWARE GraphQL APIs, or with a subset that can be deployed on-premise via a Docker container. Learn more.
- Powered by an AI Ecosystem
Leverage advanced object detection machine learning algorithms from the Veritone managed cognitive engine ecosystem — including algorithms from Veritone, niche providers, and industry giants.