Professor Kuo Delivered Keynote at ICCE-TW 2018

The IEEE International Conference on Consumer Electronics – Taiwan, 2018 was held from May 19th to 21th at National Chung Hsing University. As a grand forum for scholars and professional persons all around the world, ICCE-Taiwan aims to initiate profound discussions on research and discovery in electronics and relevant professional field.

MCL Director, Professor C.-C. Jay Kuo, gave the opening keynote at this conference. The title of his keynote speech is “RETHINKING CONVOLUTIONAL NEURAL NETWORKS (CNNS)”. In this speech, Professor Kuo first explained the reasons behind the superior performance of CNNs. Then, he presented an alternative solution, which is motivated by CNNs yet allows rigorous and transparent mathematical treatment, based on a data-driven Saak (Subspace approximation with augmented kernels) transform. His speech was well received by the audience.

After the keynote, Professor Kuo also served as a panelist in the panel discussion with a theme on “Artificial Intelligence on Consumer Electronics for Smart Future”. Besides Professor Kuo, there were two other panelists – Dr. Michael J Chang and Professor Robert Chen-Hao Chang. Dr. Michael Chang is Director General of Microsoft Artificial Intelligence Research and Development Center in Taiwan while Professor Robert Chang is a Distinguished Professor of Electrical Engineering at the National Chung Hsing University in Taiwan. Professor Robert Chang is also a Program Director of Semiconductor Manufacturing and Design for AI Edge Project under the Ministry of Science and Technology in Taiwan. They discussed the challenges faced by Taiwan IT industry in developing AI related business and products and answered questions from conference attendees.

By |May 20th, 2018|News|Comments Off on Professor Kuo Delivered Keynote at ICCE-TW 2018|
  • Permalink Gallery

    Congratulations to Siyang for Completing Her PhD Study at USC/MCL

Congratulations to Siyang for Completing Her PhD Study at USC/MCL

A MCL Ph.D student, Siyang Li, graduated from USC on May 10, 2018.

Siyang received her Bachelor of Engineering degree from The University of Hong Kong, Hong Kong, China, in June 2014. After that, she started her Ph.D study at University of Southern California, with the Annenberg Fellowship and the WiSE Fellowship. She joined Media Communications Lab (MCL) and was supervised by Prof. C.-C. Jay Kuo.

Her research interest was computer vision, with specialties in object detection and segmentation. During her Ph.D study, she collaborated with Google Research on multiple projects and had several corresponding publications. On April 26, she successfully defended her thesis titled “Object Localization with Deep Learning Techniques”.

Congratulations to Siyang and her families for completing her Ph.D study at USC!

By |May 13th, 2018|News|Comments Off on Congratulations to Siyang for Completing Her PhD Study at USC/MCL|

MCL Held the Semester-end Dinner

MCL held a semester-end dinner on Apr. 27 for the closure of Spring 2018. This is also a farewell party for MCL upcoming graduates, Siyang Li, Haiqiang Wang, and Eric Hsieh. The following is from the talk given by our graduates.

Saying: I would like to thank Prof. Kuo’s guidance during these years, and thank all support from MCL family. Studying in MCL and USC would be priceless memory in my life. My suggestion for young generations is that PhD program may be long, but starting doing research as early as possible is essentially important for the whole program. Also, balancing life and work is good for long-term productivity.

Haiqiang: I would like to thank all MCL members. Hope the junior students in MCL can keep our traditions of being friendly and helpful. We learning thinking methods and skills here, and we also learn to offer help here. This is one of the shining points of MCL family.

Eric: I really appreciate the chance here in MCL to work with different people on variant research topics, and I really enjoy participating in MCL routine work for MCL big family. Hope we can keep in touch in long term.

The semester-end dinner and farewell party will become a new MCL tradition. Wish all the best to our new graduates! Fight on!

By |May 7th, 2018|News|Comments Off on MCL Held the Semester-end Dinner|
  • Permalink Gallery

    Dr. Kuo Honored at ICASSP and USC Academic Honors Convocation

Dr. Kuo Honored at ICASSP and USC Academic Honors Convocation

commentsDr. Kuo, Director of MCL, received the Education Award from the IEEE Signal Processing Society for his contributions to signal, image and multimedia education. The award ceremony was held at the International Conference on Acoustic, Speech and Signal Processing (ICASSP), Calgary, Canada, on April 17.

Furthermore, Dr. Kuo was honored at the USC annual Academic Honors Convocation on April 24 evening for his new appointment as Distinguished Professor of Electrical Engineering and Computer Science and recognition as the recipient of the Provost’s Mentoring Award.

The USC Distinguished Professor is a designation awarded very selectively to those whose accomplishments have brought special renown to USC. The Provost’s Mentoring Award honors an individual faculty member whose investment in and generosity toward the academic and professional success of other USC faculty, postdoctoral fellows, graduate students, or undergraduate students demonstrate exemplary mentoring.

Congratulations to Director Kuo for his great accomplishments and well-deserved honors and recognitions.

By |April 30th, 2018|News|Comments Off on Dr. Kuo Honored at ICASSP and USC Academic Honors Convocation|

MCL Research Presented at ICASSP 2018

During 15-20 April, 2018, MCL member Junting Zhang presented her paper at 2018 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2018) in Calgary, Alberta, Canada. The title of the paper is “A Fully Convolutional Tri-branch Network (FCTN) for Domain Adaptation”, with Chen Liang and C.-C. Jay Kuo as co-authors. Here is an abstract of the paper:

“A domain adaptation method for urban scene segmentation is proposed in this work. We develop a fully convolutional tri-branch network, where two branches assign pseudo labels to images in the unlabeled target domain while the third branch is trained with supervision based on images in the pseudo-labeled target domain. The re-labeling and re-training processes alternate. With this design, the tri-branch network learns target-specific discriminative representations progressively and, as a result, the cross-domain capability of the segmenter improves. We evaluate the proposed network on large-scale domain adaptation experiments using both synthetic (GTA) and real (Cityscapes) images. It is shown that our solution achieves the state-of-the-art performance and it outperforms previous methods by a significant margin. ”

Congratulations to Junting for her successful presentation at ICASSP!

By |April 22nd, 2018|News|Comments Off on MCL Research Presented at ICASSP 2018|

Professor Kuo Delivered Keynote at ICDIS 2018

The first International Conference on Data Intelligence and Security (ICDIS-2018) was held at South Padre Island (SPI), Texas, USA, from April 9-10, 2018. The conference was organized by the University of Texas Rio Grande Valley (UTRGV). Key people in attendance were from the U.S. Department of Defense, the National Science Foundation, and other agencies focused on data intelligence and security.

Dr. Parwinder Grewal, UTRGV executive vice president for Research, Graduate Studies, and New Program Development, welcomed attendees to the conference in the open ceremony held on April 9 (Monday) morning. After his opening remarks, MCL Director, Professor Jay Kuo, gave the first keynote speech on the topic, “Why and Why Not Convolutional Neural Networks.” The abstract of his talk is given below.

“The superior performance of Convolutional Neural Networks (CNNs) has been demonstrated in many applications such as image classification, detection, and processing. Yet, the CNN solution has its own weaknesses such as robustness against perturbation, scalability against the class number and portability among different datasets. Furthermore, CNN’s working principle remains mysterious. In this talk, I will first explain the reasons behind the superior performance of CNNs. Then, I will present an alternative solution, which is motivated by CNNs yet allows rigorous and transparent mathematical treatment, based on a data-driven Saak (Subspace approximation with augmented kernels) transform. The kernels of the Saak transform are derived from the second-order statistics of inputs in a one-pass feedforward way. Neither data labels nor backpropagation is needed in kernel determination. The pros and cons of CNNs and multi-stage Saak transforms are compared.”

The South Padre Island ( is a tropical island located in the Mexico Gulf. It is famous for beautiful beaches, warm Gulf waters, fishing, boating, bird watching, and shopping. [...]

By |April 14th, 2018|News|Comments Off on Professor Kuo Delivered Keynote at ICDIS 2018|

MCL Research on CNN Incremental Learning

One fundamental problem of the convolutional neural network(CNN) is catastrophic forgetting, which occurs when new object classes and data are added while the original dataset is not available anymore. Training the network only using the new dataset deteriorates the performance with respect to the old dataset. To overcome this problem, we propose an expanded network architecture, called the ExpandNet, to enhance the CNN incremental learning capability. Our solution keeps filters of the original networks on one hand, yet adds additional filters to the convolutional layers as well as the fully connected layers on the other hand.

The proposed new architecture does not need any information of the original dataset, and it is trained using the new dataset only. Extensive evaluations based on the CIFAR-10 and the CIFAR-100 datasets show that the proposed method has a slower forgetting rate as compared to several existing incremental learning networks.

As a further extension, modifications such as pruning can be used to reduce the size of the proposed ExpandNet. Also, the Saak transform was recently proposed in [1]. It is worthwhile to compare the Saak-transform-based approach and the ExpandNet approach with respect to the new dataset.


[1] C-C Jay Kuo and Yueru Chen, “On data-driven saak transform,” arXiv preprint arXiv:1710.04176, 2017.


Image credits:

1. Image showing an illustration of the incremental learning problem.

2. Image showing the network architecture of the proposed ExpandNet, where new trainable filters added to the convolutional layers and FC layers are shown in orange while the original filters are shown in blue.


By Shanshan Cai, an alumna of MCL

By |April 8th, 2018|News|Comments Off on MCL Research on CNN Incremental Learning|

MCL Research on Medical Image Segmentation

Biomedical image segmentation is a technique for automatic detection of organ boundaries within an image, used to obtain diagnostic insights in the field of medicine. Segmentation results are used in applications involving organ measurements, cell detection and blood flow simulations [1].

Zebrafish is a vertebrate that have similar organs and tissues as humans, making it a valuable model for studying human genetics and disease [2]. In this project, we use the zebrafish organs dataset for the analysis and evaluation of medical image segmentation methodologies.

Presently, we have trained a CNN based segmentation model using RefineNet [3] and a Saak [4] based model on the zebrafish data. Our goal is to compare and analyze the performance of CNN with Saak segmentation model. In future, we aim to improve these results by adopting content adaptive Saak with clustering techniques and statistical analysis algorithms.

-By Shilpashree Rao and Ruiyuan Lin



[1]”Medical Image Segmentation”. [Online]. Available: [Accessed: 01- Apr- 2018].

[2]”Why use the zebrafish in research?”, 2018. [Online]. Available: [Accessed: 01- Apr- 2018].

[3] G. Lin, A. Milan, C. Shen and I. Reid, “RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017.

[4] C.-C. J. Kuo and Y. Chen, “On data-driven Saak transform,” Journal of Visual Communication and Image Representation, vol. 50, pp. 237–246, 2018.

Image credits:

Image showing the architecture of RefineNet is from [3].

Image showing the architecture of multi-stage Saak is from [4].

By |April 1st, 2018|News|Comments Off on MCL Research on Medical Image Segmentation|

MCL Research on Efficient Text Classification

A novel text data dimension reduction technique, called the tree-structured multi-linear principal component analysis (TMPCA), is proposed in this work. Being different from traditional text dimension reduction methods that deal with the word-level representation, the TMPCA technique reduces the dimension of input sequences and sentences to simplify the following text classification tasks. It is shown mathematically and experimentally that the TMPCA tool demands much lower complexity (and, hence, less computing power) than the ordinary principal component analysis (PCA).  Furthermore, it is demonstrated by experimental results that the support vector machine (SVM) method applied to the TMPCA-processed data achieves commensurable or better performance than the state-of-the-art recurrent neural network (RNN) approach.


by Yuanhang Su

By |March 26th, 2018|News, Research|Comments Off on MCL Research on Efficient Text Classification|

MCL Research Presented at WACV 2018

MCL member, Ye Wang presented Qin Huang’s paper at Winter Conference on Applications of Computer Vision (WACV) 2018, Lake Tahoe, NV/CA

The title of Qin’s paper is “Unsupervised Clustering Guided Semantic Segmentation”, with Chunyang Xia, Siyang Li, Ye Wang, Yuhang Song and C.-C. Jay Kuo as the co-authors. Here is a brief summary:

“With the development of Fully Convolutional Neural Network (FCN), there have been progressive advances in the field of semantic segmentation in recent years. The FCN-based solutions are able to summarize features across training images and generate matching templates for the desired object classes, yet they overlook intra-class difference (ICD) among multiple instances in the same class. In this work, we present a novel fine-to-coarse learning (FCL) procedure, which first guides the network with designed ‘finer’ sub-class labels, whose decisions are mapped to the original ‘coarse’ object category through end-to-end learning. A sub-class labeling strategy is designed with unsupervised clustering upon deep convolutional features, and the proposed FCL procedure enables a balance between the fine-scale (i.e. sub-class) and the coarse-scale (i.e. class) knowledge. We conduct extensive experiments on several popular datasets, including PASCAL VOC, Context, Person-Part and NYUDepth-v2 to demonstrate the advantage of learning finer sub-classes and the potential to guide the learning of deep networks with unsupervised clustering.”

Congratulations to Qin Huang for his successful presentation at WACV!

By |March 18th, 2018|News|Comments Off on MCL Research Presented at WACV 2018|