Still Image Data Querying TRIZ Cases
- Dialog-Based Image Retrieval Using Multimodal Contextual Data
A dialog-based image retrieval system integrates visual attributes with natural language feedback to model user intent.
- Multi-modal Neural Network Image Ranking via Visual and Textual Fusion
A multi-modal neural network combines visual features with textual metadata to generate accurate image rankings against search queries.
- Cross-Modality Visual Search via Learned Appearance Mapping
A processing system learns appearance mappings between RGB and infrared images to extend visual search across camera modalities.
- Molecule Metadata Extraction for Search Accuracy
A metadata database stores extracted molecule representations to enable precise document retrieval.
- Search System Entity Tag Indexing for Descriptive Queries
A search system indexes resources with entity tags to generate selectable user interface elements.
- Cross-Format Shape Embeddings for CAD Search Precision
Pre-computed embeddings resolve format conversion overhead by enabling precise similarity measurements without real-time attribute mapping.