Chanyou Hwang

Co-founder of Jaaem. Full-stack developer, system architect, researcher, and dreamer.
...

About Chanyou Hwang

Currently, I am working for Jaaem, an awesome platform for making people truly enjoy media contents. My interests are in mobile computing and platforms. I was a graduate student for 6.5 years (2012.03~2018.08), but I dropped the Ph.D. program because I disagreed with the rubbish philosophy of my (former) advisor. Still, I am interested in various research topics. I am happy to talk about new things. :)

Research

I researched with fellow students in NCLab
My Google Scholar is here

Project

I have worked on the projects listed below.

RAVEN:Perception-aware Optimization of Power Consumption for Mobile Games

FROG:Frame Rate Optimizer for Mobile Games

FROG RAVEN is the first system for optimizing rendering rate in mobile games by leveraging human visual perception. The system reduces game app's rendering rate whenever the succeeding frame is predicted to be perceptually similar enough. As a result, it optimizes the power consumption of mobile games without any modification of codes or binaries like a black magic!
Accepted at MobiCom'17 and introduced at Microsoft Research Blog!
[1-min Video], [Demo Video 1] , [MSR Blog]

TalkBetter

TalkBetter is a system for helping children with language delay. Language delay is a problem of children who do not acquire the expected level of language at their ages. It can become a lifelong risk factor if the children are not treated on right time. Through proper conversations between the children and their parents, language delay can be cured. The parents should be cautious in their conversation. TalkBetter helps parents to improve their conversation habit in daily parent-child-conversation. It intervenes in the conversations and points out the bad language habits of parents in real-time.
Accepted at CSCW'14, Best Paper Award!

SocioPhone

SocioPhone is a platform for monitoring people's conversations to support the interaction-aware applications in daily life. The system monitors conversation of people and extracts the meta-linguistic contexts that provide useful information about conversations excluding semantics which causes privacy problems. For people to use the system in daily life, the system should be energy efficient and robust. To achieve the requirements, we developed the volume-topography-based method for conversation monitoring. The method leverages multiple phones in same conversation session and makes a 'volume-topography' to detect the turn-taking which is the building block of conversation.
Accepted at MobiSys'13, Best Demo Award at HotMobile'13

Carpool Network

Carpool Network is a novel route-based carpool matching system.
In those days(2011), carpool matching services were region-based. They just matched the people who have similar destination and departure point simply. This approach makes some burden for users. They need to estimate the amount of detour distance when they are inviting or joining in a carpool. Also, it makes them to miss the chance to carpool in the middle of a route. To remedy this, we developed the system, Carpool Network.
Video here

CV

Click here to download

Contact

Feel free to email me for anything but spam!

chanyou@jaaem.com