Laboratory for Artificial Intelligence embedded in products and processes

Computer vision covers the analysis of images in 2D, 3D, and hyperspectral spectra, enabling the interpretation of scenes with different levels of detail. Various techniques such as classification, localization, detection, segmentation, and pose detection are used to extract useful information in industrial, medical, and scientific applications.

Artificial generation of data that mimics the characteristics of real data. It is used in various applications, such as training artificial intelligence models when real data is scarce or sensitive.

Hybrid Models: They combine statistical, physical, or machine learning approaches to generate synthetic data with greater realism and coherence with the original data.
GANs (Generative Adversarial Networks): These are neural networks that pit a generator against a discriminator in a competitive learning process to produce synthetic data indistinguishable from real data, used in images, audio, and other domains.

Different data models for various problems. Forecasting, Diagnosis, Prognosis, Anomaly detection, Quantum Machine Learning, Deep Learning (MLP, RNN, CNN, Transformers, LLMs, Multimodal). Algorithm explainability (XAI): LIME, SHAP, GRAD-CAM, DeepLift.

Deployment of AI algorithms following MLOps best practices.

Processing of data from various data sources with appropriate feature extraction for each type of data source. Different domains such as Signal, Text, GIS, and Image.