Artificial intelligence (AI) and deep learning startup, Talespin has announced its foray into image tagging as a service, in addition to its chatbot. Talespin enables retail and online businesses with solutions to increase customer conversion. Its technology stack uses deep learning to classify and tag product images on different attributes like colour, collar, sleeves, occasion, etc. This coupled with the chatbot helps retail brands answer three core questions for customers shopping online – “What should I buy?”, “Is this available in my size?” and “Is there anything similar to this?”.
Founded by a team of 4 co-founders in July 2016, Talespin works as a plug and play solution for businesses enabling them to leverage advanced technologies to increase customer conversion – both in-store and online. It offers in-store tablet based assisted selling solutions to retail brands across various categories such as apparel, footwear, eyewear and beauty and wellness. It is now live in stores across Delhi/ NCR, Punjab and J&K.
Commenting on the recent development, Tapan Dixit, CEO and Co-founder, Talespin said, “Image recognition and classification which forms the crux of image tagging, is one of the simpler applications of computer vision. Computer vision is the science that aims to give capabilities similar to that of the human eye to a machine or computer. It helps computers develop a high-level understanding from digital images or even videos. Bring the two – (AI powered) chat and computer vision together, you get a chat with vision. One good example of its application is a frame finder solution for eyewear stores.”
How Does It Work?
Chat with vision can be used to build an in-store frame recommended AI which will help the store associate suggest the right kind of frames for his/her customers. With the Talespin’s in-store kiosk, a store associate at an eyewear brand captures a picture of the customers’ face. Once the image is clicked it is run through a deep learning model to generate tags based on face structure and skin tone. The frame selection takes place using a digital inventory to support hard-to-fit and picky customers and prevents the loss of a sale. Basis the tags’ match, the system suggests the design type for the frame and products that fit the selection – all in a conversational format. All suggestions are based on the data in the inventory available. Once the customer selects a frame the payment can be received on the kiosk itself. Talespin also enables integration within a brand’s current loyalty program.
Talespin’s computer vision technology stack trains a series of deep learning models (one for each of a small group of attributes) and then creates a system for using those trained models to tag new images. The training process of the models can be a long task (measured in days) but once the models are trained and tuned, tagging a new image only takes seconds.
“The applications of computer vision are multiple. All current tasks performed by the human vision system can be taught to be executed by a machine algorithm. This coupled with a chat bot gives you an AI capable of processing information from images in real time and presenting it in a conversational format”, said Dixit.
The image tagging service can help brands cut down on costs that come with current manual processes of tagging as well as reduces room for error. With the right attributes available for every product, the discoverability factor goes up several notches. Talespin’s AI and deep learning technology stack at www.talespin.ai is now available to transform how brands interact with customers and vice versa.