About the Campaign
E-commerce websites have been successful long adopting recommendation system to enhance the shopping experience for their clients. With the exploding number of items in online stores, getting correct customers’ attention requires a good recommendation in place, and undoubtedly, it becomes a vital concern for retailers to be able to compete. For online fashion shopping, the common recommendation approach, which gets metrics by the likelihood of items picked up together, do not well address personal reference and make it less favorable compared to traditional fashion store where customers can freely fit and experience products before purchase.
The recent development of AI gets the headline for generating realistic style transferring images from input data. AI algorithm likes Generative Adversarial Net – a special network structure of deep learning – can grab a person or an object in one image and transfer it to a new one with new style like clothes or background applied to. This technology together with AI is profoundly pushed by the tech industry, and hitting new performance records for shorter intervals.
This style-transferring technology shows great potential for enhancing online fashion shopping. By allowing customers experience in the same way with traditional shopping where they can experience products and find the best fit. But, this pilot idea needs to convince retails for the ratio of investment cost and returned benefit and also the maturity of the technology in the form of service reliable and deploy automation.
The recent failure case of privacy protection from cloud service providers makes it more difficult for getting customers to engage in new services without providing a commitment to privacy protection which is also challenged to prove. However, the increasing computing power of edge devices like mobile or personal computer together with modern web technology make it possible for a flexible choice of keeping services on cloud or edge devices.
Our project aims to provide the online fitting solution for enhancing customers’ experience and also proposing those key values:
– Minimal Service delivery cost: Cut down the service cost by start-of-the-art AI technology powered by modern web technology.
– Privacy concern: Provable privacy protection by moving AI services to devices, an approach without keeping user information on any central database.
– Automation and reliability: Our investment in modern system building and research effort allows us to deliver services with a lower ratio of testing time versus service reliable.