Seven innovations at Princeton University, N.J., that carry the potential to benefit society and the economy, have been awarded funding to help bring their products to the global market. Three of them include or are led by scientists of Indian origin.
Princeton’s Intellectual Property Accelerator Fund provides up to $100,000 to faculty-led teams that are working on making the inventions available to the global marketplace via licensing, startups or entrepreneurial ventures. The funds enable researchers to conduct proof-of-concept studies, gather additional data, make prototypes or otherwise demonstrate the potential applications of the technology, Princeton said in a Feb. 6, press release.
This year’s winning proposals that include Indian-origin scientists, are:
An Artificial Intelligence-driven method for tracking animal motion in research, which was awarded to Mala Murthy, professor of molecular biology and the Princeton Neuroscience Institute; Joshua Shaevitz, professor of physics and the Lewis-Sigler Institute for Integrative Genomics; and Talmo Pereira, graduate student, Princeton Neuroscience Institute.
They have developed a method that uses AI to track the details of animal postures in videos, which can be used for applications to do with anything from exploring how the behaviors of flies and mice are encoded within their nervous systems, to ecological research on larger animals, the press release said.
According to Murthy et al, the new tool, LEAP Estimates Animal Pose (LEAP), can achieve nearly human-level accuracy at detecting the location of body parts of any animal species over millions of frames of video. The system labels just a few points on the animal’s body in a few frames, and the rest is done by the neural network. “Because the training of the neural network is both fast and specific to each experiment, LEAP can be used on any experimental data,” the press release noted. The funds allocated to this project will help the team improve the software for use by the broader research community.
Naveen Verma, professor of electrical engineering at Princeton, was awarded the funding for his innovation in using technology to speed up deep learning, a type of AI that offers human-level performance in a computer in several areas including vision, speech recognition, and game play. But processing large amounts of data requires lots of energy. The technology Verma’s lab has developed, uses “charge” instead of “current” thus reducing energy consumption, increasing precision and ultimately leading to far less computational noise and better scalability, the press release said.
“A prototype of the technology is already achieving record energy efficiency and speed,” Princeton noted. The new funding will help Verma and his team to identify specific applications that will enhance usability and demonstrate the capabilities of the system.
The third award to an Indian-origin scientist went to Amit Singer, professor of mathematics and the Program in Applied and Computational Mathematics, for developing software that can see extremely small things.
Singer and his team have developed software for processing the significant amounts of data generated with the Nobel-prize winning technology known as cryo-electron microscopy (cryo-EM) which provides images at the level of individual proteins.
Singer’s software, known as Algorithms for Single Particle Reconstruction, or ASPIRE, has been downloaded hundreds of times and is free for nonprofit users, the press release said.
With funds from the new award, Singer and his group will work on improving the software to make it more user-friendly, with the goal of developing a commercial version that could make structural information more readily available for drug discovery and design, Princeton said.
With recent advances in microscopes, scientists are now able to see biological structures with near-atomic resolution. The Nobel-prize winning technology known as cryo-electron microscopy (cryo-EM) provides images at the level of individual proteins and could help researchers design new and more efficacious drugs, the Princeton press release said.