Partium's search engine uniquely connects text queries and reference images, or query images with technical descriptions in your master data to find parts easily and with minimal effort
The semantic power of Partium's Search Engine
Partium's cutting-edge, proprietary, and patented machine-learning engine enables users to search for spare parts using a unique semantic image and text search.
Partium's machine learning engine translates your master data (text descriptions and reference images) and the search queries into concepts. These concepts serve as the foundation for all searches, allowing the engine to establish a connection between the search query and your master data.
As opposed to other search engines, Partium can make the connection between a text query and reference images in your catalog, or between an image query and technical descriptions in the master data. The engine also combines input image and text queries to provide the best possible results and a more precise ranking of potential matches.
For more information on semantic text search, please refer to our article on how Partium text part search works.
How does Partium's image search perform without reference images?
The engine will search parts only with the convenience of a query image that has a lot of information in a single shot. It will then suggest results where the image query is semantically related to textual information or/and available images in the master data. Finding parts without reference images works best for standard parts for which there is an abundance of data that Partium can leverage in their machine learning models.
This is very convenient for companies that do not have reference images for their spare parts available. When it comes to custom parts, the need for reference images is more relevant because those parts can only be found through visual comparison.
How does Partium's image search leverage reference images?
Apart from finding parts by semantically relating the query with part information, Partium's search engine visually compares the query image and the reference image, thereby taking the visual appearance of the object itself into account. Consequently, the search engine finds custom parts if reference images are provided.
The example below showcases how Partium image search works for parts with and without reference images:
- The user uploaded a photo of a ball valve with a red handle
- Through visual similarity between the search image and the reference image, the part "SP69" was found.
- The search engine also understood that the part in the search image is a ball valve with a lever/ handle and proposed part "RDC2423" although this part has no reference images but only textual data.
We recommend reading our article on how to take the best images for visual search to maximize performance.