Imagine seeing every move of players from a variety of perspectives, analysing their strengths and weaknesses, seeing their favourite shots and their agility levels from start to end all on your mobile phone! Sounds a little far-fetched. But with the technology being developed by the research students of International Institute of Information Technology-Hyderabad (IIIT-H), even coaches in a small town can afford to do that.
Not that such a technology is unavailable now, but is surely unaffordable.
“We propose an end-to-end framework for automatic attributes-tagging and analysis of sport videos leveraging the recent advancements in object detection, action recognition and temporal segmentation. We predict game point segments and its outcome, players’ movement as well as their strokes,” says C.V. Jawahar, professor at IIIT-H and Amazon chair professor.
Using the data, the technology can infer players’ style and strategies. “We are currently extending and deploying this framework to analyse data captured from local sports complexes in a low-cost setting, and this would help in the prolification of sports analytics-based tools in resource constrained environments like India.”
The research is supported by Amazon that has extended financial assistance to masters and doctoral students at IIIT-H through the Research Student Support Programme. “The funding can either support some students fully and other students partially. Depending on the amount of outlay, the contribution would be used towards various expenses, including tuition fee, stipend, travel grant or field study,” said a representative of Amazon. “The support programme is spread over five years and the use of funds is at the discretion of the IIIT-H.”
About 50 research scholars and faculty members have used it for their research, particularly in the area of artificial intelligence and machine learning. “Since IIIT-H doesn’t get any public funding, we take the support of corporates for taking up high-end research. Our support will not be limited to a few students due to such funding,” said IIIT-H Director P.J. Narayanan.
In fact, the Amazon funding has helped several students pursue quality research. Some of them include a project by Ph.D student Pulkit Parikh that allows social media users to determine how much importance to be attached to commercial products being promoted on their social networks. Deepa Gurmukhdas Jagyasi, another Ph.D student, is working on the development of robust reduced-complexity transceiver algorithms for millimeter-wave system for next-generation communication.