News

Congratulations to Eddy Wu for passing his defense. His thesis title is “Deep Learning Techniques for Supervised Pedestrian Detection and Critically-Supervised Object Detection” with three major topics, as follow:

In the first topic, a boosted convolutional neural network (BCNN) system is proposed to enhance the pedestrian detection performance. Being inspired by the classic boosting idea, we develop a weighted loss function that emphasizes challenging samples in training a convolutional neural network (CNN). Two types of samples are considered challenging: 1) samples with detection scores falling in the decision boundary, and 2) temporally associated samples with inconsistent scores. Finally, we train a boosted fusion layer to benefit from the integration of these two weighting schemes. We test the corresponding BCNN on the Caltech pedestrian dataset in the experiment and observe a significant performance gain over the Fast-RCNN baseline.

In the second topic, a semi-supervised learning method is proposed for pedestrian detection in a domain adaptation setup. The proposed clustered deep representation adaptation (CDRA) method uses a small amount of labeled data to train an intial detector, extracts the deep representation and, then, clusters samples based on the space spanned by the deep representation. A purity measurement mechanism is applied to each cluster to provide a confident score to the estimated class of unlabeled data. Along with a weighted training approach, the CDRA method is shown to achieve the state-of-the-art performance against some large scale datasets.

In the third topic, we propose a new framework called critically-supervised learning that mimics children learing behaviors. Several novel components are proposed to fulfill the high level concept, including negative object proposal, critical example mining, and a machine-guided labeling process based on question answering. A labeling time model is proposed to [...]

We are so happy to welcome a new Ph.D. student, Shamim Samadi, in Fall 2017. Let us hear what she said about her research and MCL.

1. Could you briefly introduce yourself and your research interests?
My name is Shamim Samadi, and I am a Ph.D. student at the MCL lab in the department of Electrical Engineering at USC. Prior to joining MCL, I was a graduate student at the ECE department in NC State University, Raleigh, NC. I received my Bachelor’s degree in Electrical Engineering from Sharif University of Technology, Tehran, Iran. I am passionate about machine learning and have recently grown a lot of interest in deep learning. My research expertise is applied machine learning and signal processing.

2. What is your impression about MCL and USC?
With a cozy beautiful campus in the heart of Los Angeles, USC is a lively environment and a top school with an outstanding Engineering program, particularly in my major. The MCL lab at USC is an excellent platform for doing original innovative research in machine learning and computer vision, with a dynamic and inspiring atmosphere.

3. What is your future expectation and plan in MCL?
My goal is to build upon my current knowledge and expertise in machine learning and be able to make a contribution to the field.

We are so happy to welcome a new Ph.D. student, Jiali Duan, in Fall 2017. Let us hear what he said about his research and MCL.

1. Could you briefly introduce yourself and your research interests?

I’m Jiali Duan from Chinese Academy of Sciences, who is passionate and enthusiastic about life and research, never cease to explore. From 2015.9-2016.3, I interned at AuthenMetric for Face Recognition and Liveness Detection. From 2014-2017, I worked closely with Prof. Stan Z, Li and published papers related to detection, face benchmark and video recognition. I also had an intern experience at Sensetime for image/video segmentation. In 2017, I was awarded with UCAS Presidential Award and Excellent Beijing Graduate. More information about me is also available at https://davidsonic.github.io/index/.

2. What is your impression about MCL and USC?

MCL is an exuberant, marvelous family, in which I have made acquaintances with people who are intelligent and hospitable. I’m also very fortunate to meet Prof. Kuo, a kind and gentle mentor, who has a strong belief and ideal about research. As for my impression about USC, I have never ceased to be awed by its aesthetic beauty and mysteriously stunning appeal. I especially enjoy the sunshine and Romanesque architecture here at LA.

3. What is your future expectation and plan in MCL?

I would like to dedicate myself to research in the next coming years, pursuing the brightest pearl on the crown of the academia. I hope that my efforts could contribute as much as possible to our lab. My future expectation about MCL is that all group members could fight together, for the prosperity of our lab.

We are so happy to welcome a new summer intern student, Yubo, in Summer 2017. Let us hear what she said about her research and MCL.

1. Could you briefly introduce yourself and your research interests?

I’m Yubo Zhang, a junior undergraduate from Department of Automation, Tsinghua University, Beijing, China. I had participated in two projects in circuits analysis and bioinformatics area respectively in my sophomore year. My research interest turns to computer vision right now. Besides, I have passion to apply computer vision as well as deep learning methods to practical areas, such as medical images and HRI problems.

2. What’s your first impression of USC and MCL?

Although USC doesn’t hold a big campus, I am impressed by its beautiful scenery and well-designed architectures. I can feel it’s a school with confidence and proud. MCL has a friendly and hospitable environment. Everyone here has a positive attitude and is willing to help. Also, it is a lab where every member is enthusiastic about his/her work. A wonderful place to do research and it’s my honor to join MCL!

3. What’s your future expectation for MCL?

It is a good chance for me to learn more about deep learning, computer vision, and above all, how to do research in this area. Hope I can contribute much to my project and to my group.