Students of School of Information Science and Technology(SIST), ShanghaiTech, presented their group work results at the final lecture of course CS172 Computer Vision. Individual students teamed up to solve real-world problems using the techniques and knowledge mastered in this course. SIST Vice Dean Yu Zhou and Executive Lijun Tu as well as numerous graduate students attended the presentation. A total of nine groups presented their findings.
The topics of the activities were determined by each group individually. These topics include:
· Anami – cartooning real life photographs and videos with demo showing our library converted to a cartoon-like library
· Chess detection using corner detection, intensity difference and Haar feature calculation enabling computer-human chess playing feasible, featuring chessboard detection and reconstruction, chess piece detection and timed based on game timer, with demo tested by Dean Zhou
· Eye focus detection with features including eye detection, pupil detection and focal point calculation using mathematical models and approximation with demo showing the focal points of the audience
· FitCraft – A game powered by body movement detection inspired by Minecraft using Xtion Pro Live, OpenNI and OpenCV with demo showing characters building architecture, fighting enemies with movement controlled by the player’s body
· Video Quality Detection using Canny Detector, Hough Transform and Gaussian Filtering aimed at improving security system intelligence with demo video shown
· Automobile License Plate Detection - able to detect license plates on vehicles in motion or at rest based on SVM and neural network
· Vintage Photo Evolution Analysis including image colorization, age prediction and the ability to predict the appearance of human after aging with applications including finding lost people, suspect identification and social interaction
· Cartoon texture – generate a tattoo-like texture on human skin in real time with user interaction with demo demonstrating its features in detail
· Game based on Kinect body motion capture with interactive features such as dynamic difficulty with demo played by students
The session was concluded by Professor Gao, instructor of this course. Gao encouraged everyone to be able to think out of the box and predict the next step to be taken in this field. He stressed the importance of solving problems using technology in an innovative way. In addition, Gao pointed out that improving technology is as important as using technology to create applications.
CS172 Computer Vision (Fall 2016) is instructed by Professor Shenghua Gao who taught topics ranging from Geometric vision to Recognition and leading into advanced topics including image segmentation, face recognition and face detection. The wide variety of topics ensured that students would be able to have a stronger background in Computer Vision and related topics to easily advance in this field.
Reporter: Xin Qin
沪公网安备 31011502006855号
