System

A system comprising:

• circuitry configured to
• receive, via a packet-switched network, sensor signal data captured by an imaging sensor of a terminal device and user condition data including preference parameter data and a numerical constraint value;
• apply a trained convolutional neural network to the sensor signal data to compute feature amount data representing spatial layout attributes, chromatic distribution information, and detected object categories;
• determine, based on the feature amount data and the user condition data, a spatial region category and a style category by computing similarity scores between the feature amount data and a plurality of stored profile vectors;
• generate a structured input sequence as a descriptive text encoding the spatial region category, the style category, the feature amount data, and the numerical constraint value;
• input the structured input sequence and the sensor signal data into a generative neural network model to generate visual representation data;
• apply an object detection model to the visual representation data to identify element regions and to generate attribute information data for each element region;
• compare the attribute information data with data records stored in a storage device to select candidate data records, and execute a combinatorial optimization process to determine a combination of the candidate data records satisfying the numerical constraint value;
• and
• transmit, via the packet-switched network, proposal data comprising the visual representation data, information of the candidate data records, and the structured input sequence to the terminal device for rendering on a display of the terminal device.

A method performed by a system comprising circuitry, the method comprising:

• receiving, via a packet-switched network, sensor signal data captured by an imaging sensor of a terminal device;
• applying a trained convolutional neural network to the sensor signal data to extract a multidimensional feature vector representing structural attributes and chromatic distribution information;
• inputting the multidimensional feature vector into a classification model to determine category label data indicating a classification of a spatial region represented in the sensor signal data;
• generating, based on the category label data, a structured input sequence for a generative neural network model, the structured input sequence encoding at least the category label data and configuration parameters;
• and
• transmitting, via the packet-switched network, output data generated by the generative neural network model in response to the structured input sequence to the terminal device for rendering on a display of the terminal device.