Thuật ngữ AI
Từ điển đầy đủ về Trí tuệ nhân tạo
Scene analysis
Algorithmic process for interpreting and understanding the overall content of an image or video sequence by identifying objects, their spatial relationships, and the overall context.
Object detection
Method for locating and identifying multiple instances of objects in an image using bounding boxes and predefined classes.
3D reconstruction
Process of creating three-dimensional models from 2D images or video streams, allowing reconstruction of the geometry and spatial structure of a scene.
Depth estimation
Technique for calculating the distance between the camera and points in the scene, essential for three-dimensional understanding of the environment.
Action recognition
Automatic identification and classification of human activities or object movements in a video sequence to understand the dynamics of the scene.
Spatial relationship detection
Analysis of relative positions between objects in a scene to understand their organization and geometric interactions.
Visual context analysis
Overall evaluation of visual elements to infer the semantic context, environment, and implicit information of the scene.
Scene mapping
Creation of structured representations of visual environments, including object localization and spatial topology for navigation or analysis.
Scene layout modeling
Reconstruction of the architectural and organizational structure of a visual space, including walls, floors, ceilings and general arrangement of elements.
Scene anomaly detection
Automatic identification of unusual elements or behaviors that deviate from expected normal patterns in a given visual context.
Multi-scale scene understanding
Hierarchical approach that simultaneously analyzes fine details and the overall structure of a scene for a complete and nuanced interpretation.
Multi-modal data fusion
Integration of information from different sensory sources (RGB, depth, infrared) to enrich the contextual understanding of the scene.
Semantic scene interpretation
Advanced analysis process that goes beyond object identification to understand the meaning, intent, and contextual implications of the scene.
Scene graph modeling
Structural representation of a scene as a graph where nodes are objects and edges represent their spatial and semantic relationships.
Camera pose estimation
Determination of the exact position and orientation of the camera relative to the scene, crucial for localization and 3D reconstruction.
Crowd analysis
Specific study of scenes with multiple individuals, including counting, tracking, and analysis of collective behaviors in public spaces.