AI-ordlista
Den kompletta ordlistan över AI
Advanced Object Detection
Cutting-edge techniques for accurately identifying and locating multiple objects in images and videos using modern architectures.
Semantic and instance segmentation
Partitioning images into pixel-level segments to understand object structure and spatial relationships.
3D Reconstruction and Stereovision
Creation of detailed three-dimensional models from 2D images or video streams using photogrammetry techniques.
Motion Analysis and Optical Flow
Study and quantification of pixel displacement between successive images to understand visual dynamics.
Biometric facial recognition
Identity identification and verification based on facial features with robust deep learning models.
Medical computer vision
Application of CV techniques to medical diagnosis and analysis of biomedical images such as MRI, CT scans, and X-rays.
Hyperspectral image processing
Analysis of images capturing hundreds of spectral bands to detect information invisible to the human eye.
Vision for autonomous vehicles
Multi-sensor perception systems for navigation and decision-making in autonomous driving.
AI Image Generation (GANs)
Creation of photorealistic and synthetic images using generative adversarial networks and other generative models.
Contextual Scene Understanding
Global interpretation of visual environments by analyzing relationships between objects and spatial context.
Embedded Vision and Edge Computing
Optimization of CV algorithms for real-time processing on resource-constrained devices.
Augmented and mixed reality
Integration and precise anchoring of virtual content in the real world through visual perception of the environment.
Human behavior analysis and pose estimation
Estimation of body poses and interpretation of human actions from video sequences.
Computer Vision for Agriculture
CV Applications for Crop Analysis, Disease Detection, and Agricultural Yield Optimization.
Transfer learning in computer vision
Techniques for adapting pre-trained models for specific tasks with limited data.