Indian Angel Network (IAN) has backed AuraML, a synthetic image data platform for computer vision and robotics, with USD 230,000 in pre-seed funding. Uday Sodhi, Neeraj Saran, and KRS Jamwal, the lead angel investors from IAN, led the round.
The Bengaluru-based startup will use the freshly raised funds to optimise its synthetic data generation engine. It has plans to hire skilled staff for its founding engineering team and wants to expand its footprints in the US and European markets. Additionally, the startup will use the capital for the official launch of its cloud platform in the coming months.
Launched in January 2023 by Ayush Sharma and Arjun Gupta, AuraML has developed a proprietary synthetic data engine. The company is working on developing an innovative technology using generative Al, which would enable people to create realistic-looking 3D worlds and images.
Ayush Sharma, CEO, AuraML, said in a media statement, "This investment will be a key accelerator for us to take the technology from a prototype to a fully functioning product. We plan to utilise these funds to hire our founding engineering team, develop core technologies and expand our global presence in the US and European markets."
"Millions of images need to be labeled by humans for training computer vision algorithms. This costs companies a lot of time and money. Additionally, the challenges related to data privacy and sharing and data collection of rare cases cannot be ruled out. AuraML, with its synthetic image data platform, is on a mission to offset all these concerns and improve the accuracy of the ML models, thereby allowing complete control over the generated dataset," said Padmaja Ruparel Co-Founder, IAN.
The problem AuraML is Solving
In 2021, there were 159 billion parcels shipped worldwide, according to Pitney Bowes. Parcel shipping is a booming business, relying on e-commerce shoppers for increased volume and revenues, and this number will only grow each year with logistics companies overwhelmed everywhere.
Computer vision and robotics is having a big impact already in solving this massive challenge by automating logistics parcel pick and sort capabilities.
However, with parcels varying in many different types of shapes and sizes with different packaging materials, parcel shipping companies face challenges everyday. To solve this, the computer vision algorithms need very large quantities of labelled training data to accurately detect all of these parcels' variations.
AuraML aims to solve this challenge by generating synthetic images with pixel perfect labels of any kind of parcel in any scenario to help improve the accuracy of the models and also tackle any pesky edge cases the models might struggle with traditionally.
Advertisements