Artificial Intelligence (AI) has expanded beyond the realms of storing and applying knowledge into intuition and inference, which have long been considered capacities unique to humans. According to AI experts, the singularity, when machines exceed human intelligence, is rapidly approaching. AI techniques have already been embedded in the fabric of our everyday lives.
(Healthcare and Law)
Intuition • Reasoning
Watson (by IBM) on the television quiz show Jeopardy
AlphaGo (by DeepMind)
DeepFace (by Facebook)
AI not only has been adopted by a variety of existing industries but also has expanded into new domains.
This trend continues to strengthen, bringing about a wave of massive change across industries.
Deep learning-based image and video analysis
CCTV and Defect detection services
Anomaly prediction and predictive maintenance
using time series analysis
Analysis and application of datasets
from cameras, Lidar sensors, etc.
Video content creation and capture for VR and
AR Broadcasting services using speech synthesis
Text and natural language analysis
and subsequent speech recognition
Financial time series data analysis
and forecasting and fraud detection
Healthcare and medical data analysis
and deep learning-based biological research
Autonomous delivery and logistics optimization
Unmanned stores and real-time marketing
AI has been adopted by a wide range of industries.
Intelligent CCTVs are increasingly adopted by local governments and public organizations. Unlike conventional CCTVs, they apply machine learning- or deep learning-based video analytics. SoyNet offers a variety of services based on object detection and pose detection.
Vision inspection through fast deep learning image analysis is essential for inspections in continuous processes that require real-time as well as intermittent processes.
When SoyNet is applied, deep learning-based fault detection using images from the production line can be processed in real time.
You can analyze the trend of various sensor data coming up from the process line and correlate the data,
and use it for the prediction of final quality items or pre-detect failure signs by similar principle.