This post is co-written with Jhanvi Shriram and Ketaki Shriram of Krikey.
ai/” target=”_blank” rel=”noopener”>Krikey artificial intelligence is revolutionizing the world of 3D animation with its innovative platform that allows anyone to generate high-quality 3D animations using only text or video inputs, with no prior animation experience required. At the heart of Krikey ai's offering is its powerful core model trained to understand human movement and translate text descriptions into realistic 3D character animations. However, building such a sophisticated artificial intelligence (ai) model requires enormous amounts of high-quality training data.
Krikey ai was faced with the daunting task of labeling a large amount of input data containing body movements with descriptive text labels. Manually labeling this dataset in-house was impractical and prohibitively expensive for the startup. But without these rich labels, its customers would have severe limitations on the animations they could generate from text inputs.
amazon SageMaker Ground Truth is an AWS managed service that makes it simple and cost-effective to obtain high-quality labeled data for machine learning (ML) models by combining ML and expert human annotations. Krikey ai used SageMaker Ground Truth to accelerate the development and deployment of its text-to-animation model. SageMaker Ground Truth provisioned and managed the labeling workforce, delivered advanced data labeling workflows, and automated workflows for human-in-the-loop tasks, enabling Krikey ai to efficiently obtain accurate labels tailored to their needs.
Deploying SageMaker Ground Truth
As a small startup working to democratize 3D animation through ai, Krikey ai was faced with the challenge of preparing a large set of labeled data to train its text-to-animation model. Manually labeling each data entry with descriptive annotations proved to be a time-consuming task and impractical to do internally on a large scale. With customer demand rapidly growing for its ai animation services, Krikey ai needed a way to quickly source high-quality tags across diverse and broad categories. Not having high-quality descriptive labels and tags would severely limit the animations your clients could generate from text input. Partnering with SageMaker Ground Truth provided the solution, allowing Krikey ai to efficiently source accurate labels tailored to their needs.
SageMaker Ground Truth lets you set up labeling workflows and use a private or vendor labeling workforce or a managed and sourced workforce, along with additional features like data labeling workflows, to accelerate and optimize further the data labeling process. Krikey ai chose to use SageMaker Ground Truth to leverage its advanced data labeling workflows and model-assisted labeling capabilities, further streamlining and optimizing its large-scale labeling process for training its ai animation models. Data was stored in amazon Simple Storage Solution (amazon S3) and AWS Key Management Service (AWS KMS) was used for data protection.
The SageMaker Ground Truth team provided a two-step solution to prepare high-quality training datasets for the Krikey ai model. First, the team developed a custom labeling interface tailored to Krikey ai’s requirements. This interface enabled annotators to deliver accurate captions while maintaining high levels of productivity. The user-friendly interface provided annotators with several options to add multiple, detailed descriptions, helping them implement comprehensive data labeling. The screenshot below shows an example.
Second, the team sourced and managed a workforce that met Krikey ai's specific requirements. Krikey ai needed to quickly process a large number of data inputs with concise and descriptive English tags, labels, and keywords. Rapid processing of the large amount of input data allowed Krikey ai to quickly enter the market with its unique 3D animation platform.
An integral part of Krikey ai's successful partnership with SageMaker Ground Truth was the ability to frequently review and refine the labeling process. Krikey ai held weekly calls to examine sample tagged content and provide feedback to the SageMaker Ground Truth team. This allowed them to continually update the guidelines for what constituted a high-quality descriptive label as they moved through different categories. Having this deep involvement and ability to recalibrate labeling criteria was critical to ensuring rich, accurate labels were captured across all of their data, something Krikey ai would not have been able to achieve on its own.
The following diagram illustrates the architecture of SageMaker Ground Truth.
General architecture
Krikey ai built its ai-powered 3D animation platform using a comprehensive set of AWS services. Essentially, they use amazon Simple Storage Solution (amazon S3) for data storage, amazon Elastic Kubernetes Service (amazon EKS) for running containerized applications, amazon Relational Database Service (amazon RDS) for databases, amazon ElastiCache for storage in-memory cache and amazon Elastic Compute Cloud (amazon EC2) instances for compute workloads. Their web application was developed using AWS Amplify. The critical component that enables their text-to-animation ai is SageMaker Ground Truth, which allows them to efficiently label a massive set of training data. This AWS infrastructure enables Krikey ai to offer its direct-to-consumer ai animation tool to customers around the world and allows enterprise customers to deploy Krikey ai's base models using amazon SageMaker JumpStart, as well as host the editor themselves. code-free 3D animation within your own AWS environment.
Results
Krikey ai's partnership with SageMaker Ground Truth allowed them to quickly build a massive set of richly labeled motion data in just 3 months and generate high-quality labels for their large data set, boosting their ai model from text to animation state-of-the-art, accelerated their time to market and saved more than $200,000 in labeling costs.
“amazon SageMaker Ground Truth has been a game changer for Krikey ai. Their skilled workforce and streamlined workflows allowed us to quickly label the huge data sets needed to train our innovative text-to-animation ai models. What would have taken our small team months, SageMaker Ground Truth helped us accomplish in weeks: accelerating our ability to bring transformative generative ai capabilities to media, entertainment, gaming, and sports. With SageMaker Ground Truth as an extension of our team, we achieved our goal of providing an easy-to-use animation tool that anyone can use to animate a 3D character. This simply would not have been possible without the speed, scale, and quality labeling that SageMaker Ground Truth offers. “They were a real force multiplier for our ai development.”
– Dr. Ketaki Shriram, Co-Founder & CTO, Krikey ai.
Conclusion
Saving time and money, along with access to premium labeled data, highlights the immense value that SageMaker Ground Truth offers to startups working with generative ai. To learn more and get started, visit amazon SageMaker Ground Truth.
About Krikey ai
Krikey's ai animation tools allow anyone to animate a 3D character in minutes. Character animations can be used in marketing, tutorials, games, films, social media, lesson plans, and more. In addition to a video-to-animation and text-to-animation ai model, Krikey offers a 3D editor that creators can use to add lip-synced dialogue, change backgrounds, facial expressions, hand gestures, camera angles, and more to their animated videos. Krikey's ai tools are available online at ai/” target=”_blank” rel=”noopener”>www.krikey.ai today on Canva Apps, Adobe Express, and AWS Marketplace.
About the authors
Jhanvi Sriram She is the CEO of Krikey, an artificial intelligence startup she co-founded with her sister. Prior to Krikey, Jhanvi worked at YouTube as a production strategist in creator community operations and programs, which sparked her interest in working with content creators. In 2014, Jhanvi and her sister, Ketaki Shriram, co-produced a feature film that premiered at the Tribeca Film Festival and was acquired by Univision. Jhanvi holds a BBA and MBA from Stanford University and an MFA in Film Production from USC.
Dr Ketaki Shriram is the CTO of Krikey, an ai animation startup. Krikey's no-code 3D editor allows anyone to create 3D content regardless of experience. Krikey tools can be used to produce content for games, movies, marketing materials, and more. Dr. Shriram received her bachelor's, master's, and PhD degrees from the Stanford Virtual Human Interaction Laboratory. She previously worked at Google (x) and Meta's Reality Labs. Dr. Shriram was selected for the Forbes 30 Under 30 2020 class in the Games category.
Amanda Lester Amanda is a Senior Commercialization Specialist at AWS, helping put ai and ML into the hands of every developer and machine learning engineer. She is a seasoned business executive with a proven track record of success at fast-growing technology companies. Amanda has extensive experience leading strategic commercialization efforts for high-growth technology. She is passionate about helping accelerate the growth of the tech community through programs to support gender equality, entrepreneurship, and STEM education.
Yulia Rizhevsky is responsible for the growth and commercialization of AWS human-in-the-loop services, serving customers to build and tune ai models. His team works with AWS customers on the cutting edge of generative ai looking to leverage human intelligence to guide models toward desired behavior. Prior to AWS, Julia developed and launched consumer products in payments and financial services.
amy dani is a Senior Technical Program Manager at AWS focusing on ai/ML services. Throughout his career, he has focused on delivering transformative software development projects for the federal government and large enterprises in industries as diverse as advertising, entertainment, and finance. Ami has experience driving business growth, implementing innovative training programs, and successfully managing complex, high-impact projects.