Developing backend service of Siemens Dashboard web application on top of Apache Solr.

Creating user interface to support exporting user’s data into Json-DL or GraphML format.

Developed a system that allows remote users to have “face-to-face” communications and immersive interactions in virtual reality environment. And accomplished my undergraduate thesis by illustrating and analyzing the VR system.

Proposed a novel augmented reality guidance modality for improving the precision of liver tumors punctures by 20%.

Analyzed in detail how reference objects influence users’ distance awareness in virtual reality environment.

Participated in developing software defect management system for IT solution and service products.

Used Myeclipse to maintain Spring MVC which serves to handle HTTP requests, connect to database to get target data through MyBatis then return customized responses.

Managed 47 members and served for over 3000 students’ social practices in our Faculty.

Accumulated planning large scale activities (over 200 participants) experience and enhanced management ability.

Acquired the ability to produce attracting slides and to give a fluent speech to over 200 audiences without draft.

Set up a Node.js server to support real-time bidirectional event-driven communication over TCP.

Wrote C# script to build multi-platform clients in Unity 3D to emit and handle events.

This network framework is deployed on both of my AR and VR projects for efficient data transmission.

Developed a system to track optical markers (attached on patient, ablation needle and AR glasses) in real-time.

Transferred transformation of optical markers to the prediction model using the robust network.

Precisely overlay the holographic display (respiration prediction model output) to the real world using transformation matrix and fast RANSAC algorithm.

Equipped with AR glasses and our system, surgeons can “see through” patients’ bodies to precisely puncture tumors.

Inherited and improved the restful user interface design and effective interaction methods of my VR meeting system.

We performed comparison experiments on more than 30 users and collected feedback with NASA-TLX. We adopted the one-way ANOVA with post hoc Tukey to analyze the error distance data.

The results show users can target spatial object efficiently (with less task load and error) in our guiding system.

Experiment: cartpole, a game to balance the pole to get higher score.

Coded to realize four human-agent interaction strategies(reward shaping, q augmentation, control sharing, action biasing) and our proposed adaptive learning strategy involving simulated oracles and real human feedback.

Conducted the experiment among 10 participants for 10 runs each strategies, collected data and quantitatively analyzed results in the paper.

Defined a programming language called sample S++ including function call, 29 reserved words(switch, while, break, et al.), compound statements, arithmetic operations and relational operations with priority, et al.

Accomplished lexical analysis, grammatical analysis, semantic analysis, intermediate codes generation and interpretive execution.

Familiar with using HTML, CSS, javascript, Bootstrap 4 to construct an elegant web font-end.

Familiar with using Docker, Kubernetes, Flask(python) Express, MongoDB, Mongoose to establish a robust back-end.