AI Glossary
The complete dictionary of Artificial Intelligence
Automated Data Augmentation
Automatic generation of image transformations (rotations, zooms, contrast modifications) to enrich the dataset and improve model robustness.
Automatic Feature Engineering
Automated creation of relevant visual features from raw images without manual intervention to optimize learning.
Automatic Model Selection
Algorithm that automatically evaluates and selects the best vision model among several candidate architectures based on performance metrics.
AutoML Image Classification
Automated system that builds, trains, and deploys models to automatically categorize images into predefined classes.
AutoML Object Detection
AutoML solution that automatically generates models capable of locating and identifying multiple objects in the same image with bounding boxes.
AutoML Semantic Segmentation
Automation of creating models that assign a class to each pixel of an image for detailed scene understanding.
AutoML Instance Segmentation
Automatic generation of models that distinguish and individually segment each object instance in an image at the pixel level.
Auto-annotation of images
Process using pre-trained models to automatically generate labels and annotations on unlabeled images.
Vision Transformers (ViT) AutoML
Automation of the architecture and training of transformer-based models specifically adapted for computer vision tasks.
Automated Model Compression
AutoML techniques (pruning, quantization) that automatically reduce the size of vision models to optimize their deployment.
Edge Deployment AutoML
Automation of the optimization and deployment of vision models on resource-constrained devices (mobile, IoT).
Zero-Shot Learning AutoML
AutoML system capable of recognizing object classes never seen during training using semantic descriptions.
Few-Shot Learning AutoML
Automation of learning vision models with very few examples per class using meta-learning techniques.
Automated Active Learning
System that intelligently selects the most informative images to be manually annotated to maximize learning efficiency.
AutoML Pipeline for Computer Vision
Complete automated pipeline of preprocessing, training, validation, and deployment steps for computer vision projects.
Automated Ensemble Learning
Automatic combination of multiple vision models to improve predictive performance through voting or stacking techniques.