News

With the advent of Web 2.0 and ubiquitous adoption of low-cost and high-resolution digital cameras, users upload and share images on a daily basis. This trend of public image distribution and access to user-friendly editing software such as Photoshop and GIMP has made image forgery a serious issue. Splicing is one of the most common types of image forgery. It manipulates images by copying a region from one image (i.e., the donor image) and pasting it onto another image (i.e., the host or spliced image). Forgers often use splicing to give a false impression that there is an additional object present in the image, or to remove an object from the image. A spliced image from the Columbia Uncompressed [1] dataset is shown above. Image splicing can potentially be used in generating false propaganda for political purposes. For example, during the 2004 US Presidential election campaign, an image that showed John Kerry and Jane Fonda speaking together at an anti-Vietnam war protest was released and circulated. It was discovered later that this was a spliced image, and was created for political purposes. The spliced image and the two corresponding authentic images can be seen above [2].

Early work on image splicing detection only deduced whether a given image has been spliced or not, and no effort to localize the spliced area was attempted. The problem of joint splicing detection and localization has only been studied in recent years. For the problem of image splicing localization, one has to determine which pixels in an image have been manipulated as a result of a splicing operation.

One of the MCL members, Ronald Salloum, is currently working on an image splicing localization research project funded by the Defense Advanced [...]

Automatic image segmentation has always been an important topic in medical imaging. Many medical applications, such as delineating heart structures, rely heavily on the accurate segmentation results. Nowadays, manual segmentation is still required in many applications. Manual segmentation is not only time-consuming and tedious but also prone to human error. One of MCL members, Ruiyuan Lin, is working on this research topic.

Many methods have been proposed to automate the segmentation process, ranging from region growing and active contour models to multi-atlas segmentation. In our research work, we focus on the convolutional neural networks (CNN) based segmentation method. We attempted several segmentation networks such as fully convolutional networks (FCN) and residual networks, compared their performance with other methods, and analyzed the strengths and problems of the networks. We are planning to further explore the use of CNN on more complicated medical images such as cross-domain images.

Image credit: both images are modified from the MRI images in the Left Atrium Segmentation Challenge dataset:
Tobon-Gomez C, Geers AJ, Peters, J, Weese J, Pinto K, Karim R, Ammar M, Daoudi A, Margeta J, Sandoval Z, Stender B, Zheng Y, Zuluaga, MA, Betancur J, Ayache N, Chikh MA, Dillenseger J-L, Kelm BM, Mahmoudi S, Ourselin S, Schlaefer A, Schaeffter T, Razavi R, Rhode KS. Benchmark for Algorithms Segmenting the Left Atrium From 3D CT and MRI Datasets. IEEE Transactions on Medical Imaging, 34(7):1460–1473, 2015.

Advanced driver assistance systems (ADAS) have attracted more and more attention nowadays, where various IT technologies are introduced to vehicles to enhance driving safety and automation.

MCL members, Junting Zhang and Yuhang Song, together with MediaTek Inc. have started a collaborative research project on ADAS-oriented deep learning technologies since January 2016. Single-image-based traffic scene segmentation and road detection have been studied extensively throughout 2016. We adapted the state-of-the-art general-purpose CNN architectures to urban scene semantic segmentation task, overcoming the cross-domain issue. On the other hand, computational and memory efficiency have always been our major concerns, we were also devoted to simplify the network structure and reduce redundant computation.

In 2017, we will explore the deep learning technologies for video processing. Although there are many interesting results in semantic urban scene understanding based on the CNN technology, semantic video understanding is still a challenging problem. We will try to find a semantic video understanding method that outperforms the single-image-based algorithms. To address this type of problems, we will exploit the temporal information.

We would like to say congratulations to five of MCL Alumni–Chen Chen, Shangwen Li, Hao Xu, Jian Li, and Xiaqing Pan–for passing their PhD defenses and starting their career lives in such great companies of the U.S.. This year, we will have Chen and Shangwen joining Facebook, Hao joining Google, Jian joining Apple, and Xiaqing joining Bloomberg. We are so glad to have them to share their experience and advises for us.

Chen Chen:
I joined MCL 6 years ago as a master student. In the first seminar talk, Prof. Kuo’s sharing totally changed my view of being a researcher. It is the moment that I made the hard decision to become a PhD student. As expected, joining MCL creates unexpected benefits to my life. The precious platform provides many opportunities for me to learn and to grow. It allows me to meet many excellent students and scholars. Working with them is like visualizing myself in a mirror, which teaches me to know myself before growing. MCL also creates huge challenges to push me to the extreme. Every exam, project due and paper submission are companied with thousands of hours of rigorous and humble team work. “Aim high and act low” is the most impressive lesson I learned from the hard process.
It is the precious experience in MCL that offers the Facebook position at the end of my PhD. However, different from what I learned in MCL, I expect to gain more working experience and entrepreneurship in Facebook. I would also like to see more students from our lab can join Facebook to strengthen the MCL alumina team in the future.

Shangwen Li:
During my past internship at Facebook, I worked on a [...]

Congratulations to Xiaqing Pan for passing his defense on January 23, 2016. His Ph.D. thesis is entitled “Machine Learning Methods for 2D/3D Shape Retrieval and Classification”.

Abstract of thesis:

Shape classification and retrieval are two important problems in both computer vision and computer graphics. A robust shape analysis contributes to many applications such as manufacture components recognition and retrieval, sketch-based shape retrieval, medical image anaysis, 3D model repository management, etc. In this dissertation, we propose three methods to address three significant problems such as 2D shape retrieval, 3D shape retrieval and 3D shape classification, respectively.

First, in the 2D shape retrieval problem, most state-of-the-art shape retrieval methods are based on local features matching and ranking. Their retrieval performance is not robust since they may retrieve globally dissimilar shapes in high ranks. To overcome this challenge, we decompose the decision process into two stages. In the first irrelevant cluster filtering (ICF) stage, we consider both global and local features and use them to predict the relevance of gallery shapes with respect to the query. Irrelevant shapes are removed from the candidate shape set. After that, a local-features-based matching and ranking (LMR) method follows in the second stage.  We apply the proposed TSR system to three shape datasets: MPEG-7, Kimia99 and Tari1000. We show that TSR outperforms all other existing methods. The robustness of TSR is demonstrated by the retrieval performance.

Second, a novel solution for the content-based 3D shape retrieval problem using an unsupervised clustering approach, which does not need any label information of 3D shapes, is presented.  The proposed shape retrieval system consists of two modules in cascade: the irrelevance filtering (IF) module and the similarity ranking (SR) module. The IF module attempts to cluster gallery shapes that [...]