Tag Archives: Computer Vision

19th BMVA Computer Vision Summer School

bmva2014.jpegBMVA Computer Vision Summer School 2014 was successfully hosted at Swansea from 30 June to 4 July 2014. A total of 67 delegates from 14 different countries ( 21 from outside UK) were attending the 19th edition of this summer school. 17 speakers from both academia and industry delivered 19 lectures and 2 lab sessions.

Click here to visit the summer school website.

RIVIC Graduate School 2014

Tim Weyrich talk at RIVIC graduate school 2014, Swansea

Tim Weyrich talk at RIVIC graduate school 2014, Swansea

The fifth annual RIVIC Graduate School was held 16th-17th June 2014 in Swansea, setting a new record of 37 talks by students and researchers, with 3 keynote talks from Majid Mirmehdi (Bristol), Tim Weyrich (UCL) and Kurt Debattista (Warwick). The event was arranged over two days in sunny Swansea with a wide range of talks in the topics of Vision, Graphics and Visualisation (Visual Computing). Our conference dinner was held in the Grape and Olive, the 27th floor of the Meridian tower offering superb views over Swansea bay, the city and local countryside.

The RIVIC Graduate School is an excellent opportunity for researchers across the RIVIC sites (Swansea, Aberystwyth, Bangor and Cardiff) to meet up, discuss research projects, seek ideas and solutions, collaborate and generally see the active and varied Visual Computing research ongoing across Wales. We also have keynote talks from excellent researchers, including this year from Majid Mirmehdi (Bristol), Tim Weyrich (UCL) and Kurt Debattista (Warwick).

Majid Mirmehdi (Bristol) spoke about the SPHERE project. In particular he gave a case study of using low cost consumer devices such as Kinect to measure progression of rehabilitation. The ultimate aim is to instrument houses with several sensors and video systems to monitor the health of occupants. He also gave some real-time automotive applications such as road sign detection and in car human/car interaction detection.

Tim Weyrich’s (UCL) talk “Bespoke Computer Graphics Systems for Cultural Heritage Applications” demonstrated how off-the-shelf devices can be re-engineered and integrated with novel software and algorithms to provide robust solutions to real-world heritage problems such as matching up smashed mosaic tiles to aid reconstruction and restoration work.

Kurt Debattista’s (Warwick) talk “High-Fidelity Graphics” covered the past, current and future challenges for high quality global illumination rendering in terms of usability, computational load and adoption. Kurt contextualised his work with industrial experience in application areas of car model rendering (with Land Rover) and architectural applications (with Arup).

Subject specific 3D human pose interaction classification

Screen Shot 2014-05-09 at 10.16.44In this work, we investigate whether it is possible to distinguish conversational interac- tions from observing human motion alone, in particular subject specific gestures in 3D. We adopt Kinect sensors to obtain 3D displacement and velocity measurements, followed by wavelet decomposition to extract low level temporal features. These features are then generalized to form a visual vocabulary that can be further generalized to a set of topics from temporal distributions of visual vocabulary. A subject specific supervised learning approach based on Random Forests is used to classify the testing sequences to seven dif- ferent conversational scenarios. These conversational scenarios concerned in this work have rather subtle differences among them. Unlike typical action or event recognition, each interaction in our case contain many instances of primitive motions and actions, many of which are shared among different conversation scenarios. That is the interactions we are concerned with are not micro or instant events, such as hugging and high-five, but rather interactions over a period of time that consists rather similar individual motions, micro actions and interactions. We believe this is among one of the first work that is devoted to subject specific conversational interaction classification using 3D pose features and to show this task is indeed possible.

pdficon_largeJ. Deng, X. Xie, and B. Daubney, A bag of words approach to subject specific 3D human pose interaction classification with random decision forests, Graphical Models, Volume 76, Issue 3, Pages 162–171, May 2014.

More details can be found at the Swansea Vision website.

Segmentation of biomedical images using shape prior

Screen Shot 2014-05-09 at 10.26.02In this article, a new level set model is proposed for the segmentation of biomedical images. The image energy of the proposed model is derived from a robust image gradient feature which gives the active contour a global representation of the geometric configuration, making it more robust in dealing with image noise, weak edges, and initial configurations. Statistical shape information is incorporated using nonparametric shape density distribution, which allows the shape model to handle relatively large shape variations. The segmentation of various shapes from both synthetic and real images depict the robustness and efficiency of the proposed method.

pdficon_largeS. Y. Yeo, X. Xie, I. Sazonov, and P. Nithiarasu, Segmentation of biomedical images using active contour model with robust image feature and shape prior, International Journal for Numerical Methods in Biomedical Engineering, Volume 30, Issue 2, pages 232–248, February 2014.

More details can be found at the Swansea Vision website.

Integrated Segmentation and Interpolation of Sparse Data

Screen Shot 2014-05-09 at 09.51.59We address the two inherently related problems of segmentation and interpolation of 3D and 4D sparse data and propose a new method to integrate these stages in a level set framework.
The interpolation process uses segmentation information rather than pixel intensities for increased robustness and accuracy. The method supports any spatial configurations of sets of 2D slices having arbitrary positions and orientations. We achieve this by introducing a new level set scheme based on the interpolation of the level set function by radial basis functions. The proposed method is validated quantitatively and/or subjec- tively on artificial data and MRI and CT scans and is compared against the traditional sequential approach, which interpolates the images first, using a state-of-the-art image interpolation method, and then segments the interpolated volume in 3D or 4D. In our experiments, the proposed framework yielded similar segmentation results to the sequential approach but provided a more robust and accurate interpolation. In particular, the interpolation was more satisfactory in cases of large gaps, due to the method taking into account the global shape of the object, and it recovered better topologies at the extremities of the shapes where the objects disappear from the image slices. As a result, the complete integrated framework provided more satisfactory shape reconstructions than the sequential approach.

pdficon_largeA. Paiement, M. Mirmehdi, X. Xie, and M. Hamilton, Integrated Segmentation and Interpolation of Sparse Data, IEEE Transactions on Image Processing (T-IP), volume 23, issue 1, pages 110-125, January 2014.

More details can be found at the Swansea Vision website.

Biometric Lab

Equipment in our biometric lab - 3D/4D capture, high speed cameras, kinect

Biometric lab

We are just setting up our new biometric lab with 3D/4D cameras from 3dMD, Point Grey Gazelle high speed cameras, Kinect and Canon EOS cameras. The equipment will be primarily used for 3D and 4D reconstruction experiments by Jason Xie and Gary Tam.

Jingjing Deng passes MSc by Research

Congratulations to Jingjing Deng who successfully defended his MSc by Research viva. He became the first MSc by Research graduate from the department since the very recent introduction of this one year research only degree scheme. His thesis is titled “Towards Human Interaction Modelling”.

Jingjing Deng was under the supervision of Dr. Xianghua Xie. The external examiner was Prof. Reyer Zwiggelaar (Aber) and the internal examiner was Dr. Gary Tam.