Abstract: A recognition method includes the following steps. A text is analyzed by a language recognition network to generate an entity feature, a relation feature and an overall feature. An input image is analyzed by an object detection network to generate candidate regions. Node features, aggregated edge features and compound features are generated by an enhanced cross-modal graph attention network according to the entity feature, the relation feature, the candidate regions and the overall feature. The entity feature and the relation feature are matched to the node features and the aggregated edge features to generate the first scores. The overall feature is matched to the compound features to generate second scores. Final scores corresponding to the candidate regions are generated according to the first scores and the second scores.

Abstract: A finger-touch tracking system is provided, including a magnet, a magnetic field sensing unit, and a computing unit. The magnet generates a magnetic field. The magnetic field sensing unit moves relative to the magnet, and has a substrate and a plurality of sensors disposed on the substrate. The sensors detect a magnetic flux density of the magnet, so as to generate a tracking signal reflecting a relative position of the magnet with respect to the magnetic field sensing unit. The computing unit receives the tracking signal via a network, and employs the received tracking signal as an input instruction upon the magnitude of the magnetic field.

Abstract: Techniques for locating information in a document relevant to an interest of a user are provided. Information defined by the user of a document browser is collected. A context model is generated using the collected information. A document selected by the user is obtained. The document is divided into one or more segments. A relevance value is computed for each of the one or more segments by comparing each of the one or more segments to the context model. The relevance value represents a relationship to an interest of the user. Each of the one or more segments with the computed relevance value is presented in a defined organizational area of a display. The one or more segments presented on the display are linked to a corresponding one or more segments in the document.

Abstract: Techniques for locating information in a document relevant to an interest of a user are provided. Information defined by the user of a document browser is collected. A context model is generated using the collected information. A document selected by the user is obtained. The document is divided into one or more segments. A relevance value is computed for each of the one or more segments by comparing each of the one or more segments to the context model. The relevance value represents a relationship to an interest of the user. Each of the one or more segments with the computed relevance value is presented in a defined organizational area of a display. The one or more segments presented on the display are linked to a corresponding one or more segments in the document.