NanoNets for Ecommerce
Extract Information from Images
Extract attributes and tags from Images. The model can be built to extract the tags that your business needs. Answer questions like what is the color of the tshirt and which is the brand of the shoe. Give a few hundered images for every tag and extract any information you want from images.
NanoNets beat human level accuracy. After receiving data we provide feedback about the accuracy of the model and compare it against your required SLAs.
Test Image Similairity
Use NanoNets to find images that are similar. Similarity is set to be whatever you want. You can find similar patterns, similar colors or similar articles. Pass two images and test how similar they are. NanoNets returns a score describing how similar two images are. NanoNets are trained to understand similarity by learning from data you provide. Pass the model a few images of images that are similar and those that are dissimilar, it learns to identify similarity from this.
Recommend Products
Use NanoNets to recommend similar products. Recommendation can be based on products that are visually similar or products that would look good together or products that would complete an outfit. Train the model by providing it a list of all your products and some examples of product recommendations. Example these jeans look good with this tshirt or since you bought these jeans, you might also like these jeans.