Learning a new language by leveraging a UGC-recommending system of multimedia content makes it easy and fun.

By empowering users to upload text mnemonic rules, interesting images and video clips, we are eventually creating a community for people who are struggling with language learning.

It's a more intuitive way to interact with people at social events by adding their digital social profile with real person in space.

By linking personal mobile device with the system, editing digital profiles through personal mobile devices will be reflected in mixed reality in real time.

It creates a new social network experience which real time digital information supplements the previous unknown information to entities in the mixed spaces.

This personal assistant is designed to provide people with a more visual and vivid guidance to help with their personal habit formation. I believe this can be a better solution than voice assisant from a user interaction point of view. It leverages the immersive interaction experience that AR provides to help people in a more convincing way and a more effective behavior intervention.

The project aimed to create an experience which transfers events between reality and virtual space. Actions that are initiated from reality to virtual world will create real world effects from virtual space afterwards. It's no longer a single direction control always originated from reality anymore. Just like the Turning Test is for AI, the ultimate test for mixed reality will be whether we can distinguish which pieces are real which are virtual when seeing dominoes' falling one pushing another.

An augmented reality application that helps students to learn language.

The application is able to take voice input from teachers who setup the language learning goals by looking at the real world objects and binding pronunciations with them. After the setup stage, the students will try to look for these objects in the real world by listening and searching the spatial sounds source as guidance. It is an effective way of learning language vocabulary by triggering a genuine 'discovering' mindset by leveraging the mixed reality and virtual spacial sounds.

Robots that are capable of 'sensing' the patients' homes will be deployed to their houses after they are discharged from hospitals for follow-up medical care.

The robots' 'sensing' system includes a map-generation system that can create maps of the patients' homes which enable a better observation for hospitals. The robots can also 'sense' the patients by a human-tracking system.

Through the system, doctors are able to observe multiple patients who are at homes remotely without leaving hospitals and provide help in time.

The project's goal is trying to replace traditional hand control interfaces such as joy sticks, keyboards so people with no limb mobility can be better facilitated.

The control interface module can take in multiple control signals through facial expressions via Emotive headset toolkit. The outputs are processed from the raw signals to filtered results.

The modularized interface is adapted to control wheelchair's navigation system so people who can't use their limbs(locked in syndrome) are still able to have certain mobility. The wheelchair users are able to navigate through the new interface in relatively complex environments. By recording the manipulation data with wheelchair's trajectory, certain optimizations were explored by utilizing these history data.

It can also be adapted to personal computers for a web browsing experience. It provides certain convenience to people who can't use mice and keyboards. Users are able to browse various predefined websites with certain navigation abilities.

The same controller can also be easily adapted to become a control interface for robotic systems.

A different approach to replace the physical control interfaces (Joystick, keyboard, mouse).

I worked on the three-door task as part of the DARPA challenge 2013-2014. It requires the robot to locate three doors and open them, then go through each one.

The task used point cloud data collected through a rotating 360-degree LIDAR sensor. The robot was able to identify the doors and their handles' positions. Then the human operators are able to control the robot to open and walk across the doors.

A mixed sensor system combining IR sensors and Ultrasonic sensors is built in a highly configurable design that can do a 360-degree range of detection. This is designed as an alternative of an expensive 360-degree LIDAR to lower the cost from $3000 to less than $300.

The goal is to use real world information(personal voices and geo-locations) to augment the traditional text based post sharing system to encourage people to interact in a more personal way.

An opinion sharing platform is designed for this purpose to help people to post their thoughts on Shakespeare's literature works.

The platform takes audio recordings uploaded by people combined with the traditional text posts. It also helps people to show their geo-locations pinned on to a world map so it makes them know more about other peers and feel more comfortable sharing their own thoughts.