Ignacio Palos-ReynosoHumberto HÃjar2026-03-222026-03-22202610.1016/j.commatsci.2026.114508https://doi.org/10.1016/j.commatsci.2026.114508https://andeanlibrary.org/handle/123456789/79291enTopological defectMesoscopic physicsLiquid crystalTopology (electrical circuits)Deep learningConvolutional neural networkArtificial neural networkArtificial intelligencePhysicsComputer scienceDeep learning detection of topological defects in confined two-dimensional nematicsarticle