Developing an AI-based image editing solution for one of India’s largest online jewelry marketplaces- eJohri
About the Client
eJohri is India’s first omnichannel jewelry marketplace. The platform functions as an eCommerce solution for people to buy jewelry online and discover offline retailers nearby to make an offline purchase. It has jewelers from metropolitan, urban, and suburban cities.
The jewelry aggregator offers a diverse range of jewelry including gold, silver, kundan, platinum, zirconia, solitaire, etc. The average ticket size for online sales is INR 50,000 through the platform.
- 98% increase in efficiency
- >95% accuracy rate
- 100,000+ images analysed & corrected
For jewelers to register to the eJohri platform and start selling, they need to create a portfolio of their products. This portfolio, in the form of jewelry imagery, is sent to the eJohri team where is refined using photo-editing tools. This process included removing background, shadows, or any unwanted items in the images for a better presentation of products on the platform.
On average, this practice would take 2-3 hours for a single image to edit, which slowed down the jeweler’s onboarding to the platform. As the business expanded, it became inevitable for team eJohri to automate this image editing cycle so that jewelers can quickly onboard and start selling on the platform.
Hence, eJohri was on the lookout for a technology partner who could develop an AI-based image editing solution to accelerate the onboarding process. Some of the key expectations of the eJohri team were to
- Evaluate the most suitable, AI technology to train machine learning models and edit images in bulk
- Augment the existing data set of images to ensure a high accuracy rate of the ML model
- Ensure that the quality of the image is exactly the same as shared by the jeweller in his portfolio of products
Team Daffodil analyzed the existing business challenges and as a solution, tried computer vision techniques to remove the background from the images. Edge detection, image thresholding, unsupervised learning (K-clustering), etc. were some of the techniques that were examined individually and in combination to resolve the problem. However, these techniques worked differently on different types of jewelry and posed accuracy concerns. For that reason, computer vision techniques were dropped and machine learning models were chosen as the solution to resolve the problem.
One of the prime challenges in building the ML model was the limited data set. To augment the data set and train the model for maximum accuracy, the transfer learning technique was utilized. This allowed team Daffodil to train the model on a generalized data set of jewelry.
This fundamental model training overcame the limitations of the finite data set available. The machine learning model thus created had an accuracy rate of 95% and saved several man-hours for the eJohri team in the image editing process.
The machine learning model allowed the eJohri team to increase the efficiency in image editing by 98%. An image that would otherwise take 3-4 hours to edit in photoshop could now be edited using the ML model in just 2-3 seconds.
Read Related Case Studies
Get in Touch
Sign up for a 30 min no-obligation strategic session with us
Let us understand your business objectives, set up initial milestones, and plan your software project.
At the end of this 30 min session, walk out with:
- Validation of your project idea/ scope of your project
- Actionable insights on which technology would suit your requirements
- Industry specific best practices that can be applied to your project
- Implementation and engagement plan of action
- Ballpark estimate and time-frame for development