Gcore has introduced a cutting-edge AI Content Moderation system to automatically filter out offensive, violent, or illegal content from user-generated material. This innovative solution utilises advanced AI models in computer vision, OCR, and speech recognition, and operates seamlessly through Gcore’s extensive global network infrastructure.
Gcore Launches Real-Time AI Content Moderation Solution
Gcore, a global provider of edge AI, cloud, network, and security solutions, has launched a real-time AI Content Moderation system aimed at automating the moderation of user-generated content (UGC) to protect viewers from offensive, violent, illegal, or age-inappropriate material. This system integrates seamlessly into existing infrastructures via API and operates through Gcore’s network of over 180 edge points with a total capacity exceeding 200 Tbps.
Key Features:
- Computer Vision: Uses advanced models for object detection, segmentation, and classification to flag inappropriate visual content.
- Optical Character Recognition (OCR): Converts visible text in videos into machine-readable format for textual content moderation.
- Speech Recognition: Analyzes audio tracks to detect and flag foul language or hateful speech.
- Multiple-Model Output Aggregation: Integrates outputs from various models to make precise moderation decisions, particularly for complex issues like child exploitation.
Statement by Alexey Petrovskikh:
“Human-only content moderation has become an unmanageable task due to high volumes and costs. Gcore AI Content Moderation allows organizations to scale globally while ensuring regulatory compliance and preserving their reputation,” said Alexey Petrovskikh, Head of Video Streaming at Gcore.
Gcore, headquartered in Luxembourg, employs 600 staff across ten offices worldwide and serves multiple industries including social media, gaming, ecommerce, and education. The company’s infrastructure spans six continents, offering one of the best network performances in Europe, Africa, and LATAM with an average response time of 30 milliseconds globally.
