- Services
Design & Development
- Industries
- Generative AI
- Case Studies
- Learn AI
- Careers
- Company
Sportlight
Sports
Computer Vision / AI
Develop an action temporal localization system for detecting and analyzing specific actions, starting with "Cricket Bowling," with potential expansion to other sports.
Create a deep learning-based tool for precise temporal action detection in sports, providing real-time insights for trainers and stakeholders.
Developed end-to-end pipelines for data preparation, training, and inference, focused on action localization using models such as Boundary Sensitive Networks (BSN) and mmaction2.
Delivered a high-accuracy system for detecting cricket bowling actions, laying the foundation for expanding to other sports, with the potential to be sold as a B2B tool for trainers.
We utilized BSN (Boundary Sensitive Network) and mmaction2 to accurately localize the start and end of actions within game footage, ensuring precise detection of the "Cricket Bowling" motion.
Comprehensive pipelines were developed for data preparation and model training, ensuring that the system was robust and could generalize well to unseen footage.
The system supports local inference, allowing real-time action detection and feedback, which is crucial for live game analysis.
The architecture was designed to be extendable, enabling the future training of models for other sports like football, tennis, and more.
Accurate, real-time detection of specific actions in cricket footage.
Complete data preparation, model training, and inference pipelines.
It puts out accurate job specifications and leverages AI to produce job postings that reach the suitable candidate.
I searches career portals and social media for top candidates, and automatically matches resumes to job descriptions, ensuring only the most suitable candidates are shortlisted
The system can be adapted to other sports, making it versatile for various sports applications.
Local inference capabilities provide live analysis, improving training and coaching decisions.
We love to hear from you. Drop us a line, or give us a heads up if you’re interested in visiting us.
We love to hear from you. Drop us a line, or give us a heads up if you’re interested in visiting us.