The talk will cover how industrial safety monitoring can be automated by doing video analytics on CCTV footage. A brief introduction to deep learning and machine learning will be followed by presenting the challenges involved in detecting personal protective equipment from CCTV footage in industrial sites. Various techniques and approaches used to improve detection accuracy will also be discussed. Finally, the performance of a trained model on test videos will be presented, and several use-cases will be discussed.
The following will be the key takeaways of this talk: –
Understand deep learning for object detection
A guide on the steps involved in setting up a deep learning model for a practical problem
Learnings on industrial scenarios where AI/ML can bring in accuracy and efficiency
Sri Vallabha has expertise in machine learning, statistical analysis, predictions, reduced order modelling and simulation of physical systems. He is currently working on object recognition in image/video using deep learning with techniques like transfer learning, ensemble modelling, and fine-tuning. He has developed numerous regression and classification models for forecasting and prediction in various business scenarios. Sri Vallabha holds a PhD from IISc, Bangalore and B.Tech from IIT Madras.