Hasty Raises $3.7M in Funding to Accelerate Computer Vision AI Education


Artificial intelligence startup Hasty Ltd announced today that it has received $3.7 million in seed funding to accelerate development of its next-generation computer vision annotation tool.

Shasta Ventures led the seed funding round along with participation from coparion and iRobot Ventures.

Hasty has developed an annotation tool used to train visual AI models used by computers to understand what they see when presented with an image or video. In order for a computer to know what it is seeing, the content of the image or video must be tagged so that it can, for example, “see” different objects or human faces and read text.

However, data labeling is a tedious task hampered by the fact that most visual AI models do not train and update themselves during this process. In fact, Hasty says that around 80% of a data scientist’s time is spent labeling the data they use to train their neural networks. Because of this lag between labeling and model training, machine learning engineers often have to wait months before they can see if their annotation strategy actually works.

Hasty says its next-generation annotation tool speeds up Vision AI training because it can train models while the data is being annotated. Additionally, the more the tool is used, the faster it becomes at labeling objects, as the model helps create a cycle of automation by informing the labeling process once it has started learning.

Hasty said its AI assistants are able to predict what labels should be after annotating just a few images, allowing data scientists creating their datasets to work much faster. The tool also has an Error Finder & Manual Review bot that helps in detecting annotation errors due to misclassification, artifacts, missing objects and poor segmentation.

Meanwhile, the Rapid Feedback Loop bot offers short feedback cycles, allowing data scientists to iterate faster, validating and adjusting their data labeling approach as needed. This means they can test different approaches to ensure they are creating the best data to train each model.

“We bring our customers to the forefront and improve their ability to bring more successful vision AI projects to market faster,” said Tristan Rouillard, co-founder and chief executive of Hasty.

Hasty believes there will be a great demand for tools that can speed up the vision AI training process given the discipline’s tremendous potential across so many industries. Some examples are helping doctors diagnose diseases by viewing x-rays or scans faster, or tracking packages through supply chains.

“Vision AI has the power to unlock the future of automation and accelerate the development of new products and services in ways not seen since the web revolution,” said Isaac Roth, partner at Shasta Ventures. “Hasty’s engineering team understands this and has delivered a best-in-class solution that gets faster and smarter the more it’s used.”

Image: Hastily

Show your support for our mission by joining our community of experts, Cube Club and Cube Event. Join the community that includes Andy Jassy, ​​CEO of Amazon Web Services and Amazon.com, Michael Dell, Founder and CEO of Dell Technologies, Pat Gelsinger, CEO of Intel, and many more luminaries and experts.

Previous Why Black-run businesses have been hit hardest by the pandemic
Next Surf clubs on day four helping out with flood rescue