Imagga announced the beta release of its Visual Similarity Search API, enabling companies to implement image-based search into their software systems and applications to improve product discoverability and decrease shopping cart abandonment. Visual Search expands the core capabilities of Imagga’s powerful Image Recognition API, helping companies to maximize the searchable potential of their visual data.
“The visual search technology represents an exciting opportunity for fashion, home décor and furniture online retailers helping them to improve product discoverability, increases conversions, and decreases shopping cart abandonment while also offering rich media experience to users,” said Georgi Kadrev, Co-Founder of Imagga. “Imagga visual similarity search technology can be applied to any visual dataset regardless of the industry or the use case.”
Search by Image – Visual Search makes product discoverability easier by enabling users to search for similar items within companies’ product inventory by uploading an image from the web or one they took themselves.
Product Recommendations – The functionality is often used for the display of similar products offering bigger choice or alternatives to products out of stock.
High Visual Search Accuracy – To achieve the highest precision rate, Imagga’s expert machine learning team helps customers customize the features relevant to their image database.
Scalability – Imagga Visual Search API features a scalable infrastructure capable of serving large amounts of requests within extensive search pools.
How Visual Similarity Search Works – The technology is based on deep learning and extraction of visual and semantic features during categorization that is used as matching criteria during the search. After extracting metadata from the image, it digs through the index of images and retrieves the best results based on semantics, color, category or functional similarity.