Hello Makers!

Join us this evening to discuss how computer vision can help detect anomalies in public surveillance feeds!

Following is a brief agenda for the evening:

6 - 6:30 PM - Doors open & pizza
6:30 - 7:15 PM - Utkarsh's Talk
7:15 - 7:45 PM - Q&A & Networking
7:45 - 8 PM - Contest for special prize!

There has been significant progress in the field of Computer vision for tasks such as image classification & object detection, which have outperformed humans in various commercial & industrial applications. Nevertheless there still remains much work to do in more complex tasks such as finding anomalous events in videos especially in case of public surveillance feeds, which are extremely challenging due to factors such as occlusion, cluster , low quality feeds, etc. In this paper we will evaluate such challenges & propose a solution based on neural architectures ranging from supervised to reinforcement based approaches. Our hope is that this work is foundational to further research & study to be conducted in this area.