Precise, scalable data annotation is helping drive forward development in AI across many industries. In the interior design and furniture business, computer vision based AI applications are already having a significant impact on the way consumers shop and buy. Furniture products have been slower to shift to online shopping. This is in part due to the size and delivery costs associated with larger items, but it is also connected to consumer’s desire to see furniture in person in order to get a good idea of how it will look in their home. However, this is changing as younger millennials choose to buy furniture online. Interior design and furniture companies are keen to provide this growing share of online customers with a way of visualizing how different products will fit in their spaces.
Three dimensional room planners and augmented reality mobile apps are providing this next level shopping experience. This blog will focus on some of these recent developments and show how image annotation is central to the success of these applications.
Finding the right furniture items
Large online retailers, such as Wallmart and Ikea, have tens of thousands of furniture and interior design products that consumers must navigate through. Inventorying and categorizing all of these varied items for online shopping is an enormous task, and mistakes can lead to products being missed by consumers and the user experience suffering. Computer vision algorithms can remedy this situation by automatically identifying and grouping together objects in a large catalogue. These similar items can then be displayed to customers or can be searched for.
In order for products to be accurately identified and categorized machine learning programs must be trained with large quantities of labeled image data. The size of the databases required for precisely functioning image recognition systems is significant. Professional annotation services, like Keymakr, are able to take on the burden of dataset creation and annotation, enabling developers to concentrate on the key goals of delivering for customers.
Creating a virtual living space
A number of interior design companies are now offering customers the opportunity to design and decorate a three dimensional room. This space can be filled with products from the company’s range, allowing consumers to get a real sense of how particular pieces of furniture or decoration will fit together in the real world. Exciting research is now taking place utilising machine learning to convert two dimensional images into 3D objects. This will allow for even greater flexibility in 3d room planners with online shoppers able to take any furniture product and project it into their rendered 3D space.
Quality training data is a key part of this cutting edge computer vision technology. Collaboration between machine learning experts and image annotation specialists can streamline development and research. Access to bespoke annotation tools and experienced annotators can provide innovators with a valuable edge.
Shopping with augmented reality
Wayfair is just one of the companies providing online furniture browsers with augmented reality mobile apps. These programs allow consumers to see how a particular piece of furniture will look in their own room. Position sensing data in the phone is combined with computer vision algorithms to project furniture onto flat surfaces so that it appears proportionate and accurately placed in a given space.
Accurate object labeling is a crucial factor in the development of AR interior design use cases. Complex objects like tables, sofas, and chairs need to be carefully annotated using techniques such as polygon annotation. Image annotation services offer a full suite of annotation methods and are flexible enough to meet the demands of any project.
Streamline your machine learning project with Keymakr. Take advantage of cutting-edge annotation tools and an experienced, in-house team of annotators. Contact a team member to book your personalized demo today.