VreVRE is a real-time rendering engine that I developed with team members at Wanyi, based on Rust and WGPU. It not only has realistic rendering effects but also supports the creation of large scenes. With fast operation speed and excellent performance, it serves as the cornerstone of the entire Wanyi Dougong project.
Click here to see more. |
DSST Object TrackingDSST algorithm propose an efficient method for estimating the target scale by training a classifier on a scale pyramid. This allows us to independently estimate the target scale after the optimal translation is found. It improves the accuracy over an exhaustive scale space search method, while running at 25 times faster frame-rate.
Check here to see more. |
JDA face detection and alignmentJoint Cascade Face Detection and Alignment is an algorithm which can get a state of the art face detection algorithm. The advantage of this algorithm is that it can detect the face area and keypoint at the same time. Beside the speed is quit fast (more than real-time).
Click here to see more. |
DeepID Face RecognitionDeep Learning is a hot topic in CV area now, we implemented the famous DeepID algorithm.
Click here to see more. |
Multi-View 3D Object ReconstructionReconstruct an object into a 3D model with several pictures with different views.Use SIFT algorithm to extract the image feature points and match them with each other. Calculate the camera matrix and the 3D coordinates. Reform the point cloud into mesh with the images as texture mapping to get the final 3D model.
Click here to see more. |
Face Detection and Face RecognitionApply Haar classifier and Fisher algorithm to do the face detection and face recognition. Will change to Deep Learning later to get a better result.
Click here to see more. |
Object Tracking On CellphoneAn online open-source algorithm which is used for object tracking. It learns from the movement of the object and detects the object even if it's blocked.
Click here to see more. |
Fast animation rendering
I use maya and kinect to connect the two devices together and sending data from the kinect(client) to the maya(host) to drive a model to move. Click here to see more. |
New Oil Painting Effect |
Beyond The StageOur mission is to use interactive media to bring the magic of theater to the audience online, hoping to draw that audience to the theater once again.
Click here to see more. |
Oceanus 3Oceanus 3 is continuing the Entertainment Technology Center’s development of two interactive crowd experiences for the 2012 World Expo in Yeosu Korea. The team is based at the ETC’s Osaka Japan campus and our advisor is Dr.Salvador Barrera.
Click here to see more |
Hand AnimationCombining three devices and the same time and use them to create a new platform to control the hand model and move around, rotate in the virtual world. Plus using the HMD to see the world in 3D mode.
Click here to see more. |
SpeakEasyThe goal of the Speakeasy project is to develop a fun, effective, and compelling method to learn new languages using the Minspeak symbolic icon system.
Minspeak is a dedicated hardware/software platform that allows people to communicate by representing words and concepts with sequences of icons. This was originally developed to help the communication-disabled speak more efficiently, and we are expanding its scope to include language learning on a tablet device. Click here to see more. |
Computer GraphicsThis course provides a comprehensive introduction to computer graphics modeling, animation, and rendering. Topics covered include basic image processing, geometric transformations, geometric modeling of curves and surfaces, animation, 3-D viewing, visibility algorithms, shading, and ray tracing.
Click here to see more. |
BVWAll projects I do in last semester at ETC were called BVW, which is work together with other team members to create highly interactively game with different platforms in two weeks usually.
Click here to see more. |
Personal workHere are some of my old projects, some are for fun, some are research of fun.
Click here to see more. |