Evaluation & assessment
Evaluation & assessment · Concept diagram
Dimensions and workflow of model evaluation
MEDIA LIBRARY · SEE & UNDERSTAND
Each theme has space for a diagram, an animation and a video. Media will be added gradually to explain specific concepts and standards.
EXISTING INTERVIEW
ITU expert interview: Question 5 Rapporteur Yuntao Wang introduces AI-enabled multimedia applications. Linked from the ITU Vision page.
Watch on YouTube ↗Evaluation & assessment
Dimensions and workflow of model evaluation
Evaluation & assessment
How inputs, tasks, metrics and evaluation results connect
Evaluation & assessment
Model evaluation methods and practical testing scenarios
Platforms & engineering
How platform capabilities relate to lifecycle stages
Platforms & engineering
The model lifecycle from data preparation to deployment and operation
Platforms & engineering
Platform engineering and device-edge-cloud application examples
AI agents
How perception, memory, planning and execution connect
AI agents
How agents execute and coordinate a task
AI agents
Agent systems and typical application examples
Hardware & computing power
How chips, clusters and benchmark targets relate
Hardware & computing power
A testing workflow connecting workloads, measurements and results
Hardware & computing power
Hardware computing and energy-efficiency testing scenarios
Embodied AI
The feedback loop between perception, decisions, actions and the environment
Embodied AI
A robot's observation, action and feedback loop
Embodied AI
Embodied AI applications in real-world settings
Generation & media processing
How generation applications connect to external knowledge and media processing
Generation & media processing
A workflow for retrieving evidence, enriching context and generating answers
Generation & media processing
Generative multimedia and RAG application examples
Digital human systems
The components of appearance, voice, motion and platform capabilities
Digital human systems
Interactions with synchronized speech, expressions and motion
Digital human systems
Digital human systems and service scenarios
Federated & distributed AI
Architectural relationships between participants, coordinators, devices, edge and cloud
Federated & distributed AI
How data locations and model updates relate in collaborative learning
Federated & distributed AI
Federated learning and distributed application examples
Industry & knowledge engineering
How industry scenarios, technical capabilities and relevant standards connect
Industry & knowledge engineering
A workflow from organizing domain knowledge to application services
Industry & knowledge engineering
Application examples in industry, customer service and intelligent spaces