ViSenze’s visual search and recognition technology combines machine intelligence and computer vision algorithms to generate image and video recognition capability to process, analyse, understand and interpret any form of visual content. (Image: ViSenze Pte Ltd)
You saw a dress on a passer-by, but you are not sure of its name or brand. You tried searching for the dress online using keywords to describe it and your results turn out negative. You give up looking for the dress and the e-retailer loses a potential customer.
Local technology startup, ViSenze, has developed a set of intelligent visual search and recognition algorithms that are able to identify, tag and describe images and video content. Their visual technology is based on deep learning and computer vision that is fully-automated and scalable. It can interpret packaging such as labels and logos, drawings and sketches, patterns in fabrics and art, as well as fashion items such as bags or dresses.
Users upload an image and ViSenze’s technology will search through online databases for exact or similar matches.
Simplifying the visual web with a touch of the mobile phone camera button. View this video to gain a better understanding of the features of its image search technology
Text search versus image search
To illustrate the efficacy of its visual technology, ViSenze asked few participants at CommunicAsia 2015 to search a dress from an online photo via text search. Ninety-five per cent of them gave up their search in 90 seconds. Using ViSenze’s visual search tool, the dress was located in an online store within three seconds.
ViSenze “Find the Dress” Challenge at CommunicAsia 2015: Keyword search versus visual search
Applications in e-commerce and online advertising
ViSenze’s visual search and recognition technology is used in e-commerce, database management, and mobile applications. E-retailers tap it to help their customers search and shop without keyword guessing. The company counts international e-commerce companies like Rakuten, Zalora, Flipkart, Lazada Goodrich Global and Clozette amongst its international clientele, which also include customers from China, Japan, Taiwan, India, Southeast Asia and North America.
In online advertising, advertisers identify and tag visual content for ad targeting purposes. The visual recognition technology also filters unsafe or inappropriate images to ensure brand safe advertising. In big data analytics, it helps brands to understand visual social communications (for example, the number of re-pinned images on social media platform Pinterest) that traditional text-data engines cannot interpret.
About ViSenze Pte Ltd
ViSenze Pte Ltd, founded by computer scientists and web entrepreneurs, is a spin-off from the NUS-Tsinghua Extreme Search Centre (NExT) jointly set up by NUS and Tsinghua University of China. The research centre is supported by the Interactive Digital Media Programme Office funded by the National Research Foundation Singapore.
The company was selected as a winner for Red Herring 2014 Top 100 Global award, which honours the year’s most audacious and far-reaching private technology companies and entrepreneurs from across the globe. This is in addition to its place in the 2014 Red Herring Top 100 Asia list of Asia’s best startups.
ViSenze was named a winner for the Category of Most Innovative Infocomm Product/Solution at the 2014 National Infocomm Award ceremony. The biennial award, jointly organised by the Infocomm Development Authority of Singapore (IDA) and the Singapore infocomm Technology Federation (SiTF), is Singapore's highest accolade for infocomm innovation.
The company received US$3.5 million follow-on funding from Rakuten Ventures, Walden International and UOB Venture Management in February 2014.