Deep learning use cases for ASEAN

Deep learning is going to be the foundation of many disruptive new applications during 2017 and beyond.


As a hobby photographer, I appreciate Deep Learning features that came to action recently. I used to sit and add keywords to photos for many hours so that I could query my pictures and find pictures of interest. By adding keywords, I could go beyond EXIF data and look for objects, moods and colors I wanted to see in my photo collection. Not too long ago, these search capabilities are ‘instant’ through on-line cloud services such as Google Photos. I can query for blue skies, bright smiles …etc. after uploading my photos to such a service in the cloud. This capability is evolving and has its limitations (for e.g. I can’t search specific objects such as a double delight rose, a cricket ball…etc.) but this is a great time saver! Limitations just mean more training is needed to identify specific objects. The same technology can be applied for more complex problems in multiple domains.

Deep learning is going to be the foundation of many disruptive new applications during 2017 and beyond. As the technology evolves, the right mix of emerging technologies and trends such as IoT, Drones, robotics together with Deep Learning can bring about new ways of solving complex problems.

I want to explore some of the deep learning use cases for this region that can create impact. As a fast-growing region in Asia-Pacific, ASEAN has great potential for deep learning use cases. I have picked 3 areas but keep your thoughts coming and share your views.

1. Less time on the road:

All the cities in ASEAN face traffic jams almost every day. Metro Manila and Jakarta are two of the cities with worst traffic jams in the whole world. The problems faced includes increase in pollution, road rage, unproductive hours and waste of fuel …etc.

Various use cases can help to create innovative solutions:

a. Dedicated traffic lanes are often used in ASEAN countries but they don’t have any intelligence built into them or any decision-making capabilities. Deep Learning can be used to enhance traffic-related decisions.

b. A study from China published in the IEEE/CAA Journal of Automatica Sinica suggests that Deep Learning can be deployed to plan traffic signal timings and in turn expected to reduce wait times and make traffic queues shorte.

2. Safety and Security for all:

Security is a topic of focus around the globe due to terrorism, cyber-attacks. Some of the use cases for enterprise solutions are:

a. Deep learning together with video analytics can play a part for monitoring /identifying suspicious and abnormal activities and notification of alerts. Such a system can be trained for specific domains such as public events, airports, schools, shopping malls …etc.

1  2  Next Page 

SPONSORED LINKS