We proposed a general framework which applies classifiers with different complexity to discriminate segments in an image.
Our unsupervised
hierarchical segmentation results achieve similar or better performance in several standard benchmarks compared with the current state-of-the-art methods based on learning, and has been accepted to ACCV 2014. 
(Paper, Poster)

Hierarchical Image Segmentation without Training

We proposed an efficient algorithm which decomposes the unsupervised Multiple Foreground Co-segmentation problem into three sub-problems: segmentation, matching and figure-ground classification.

Our method improves the accuracy of the state-of-the-art method by 13% in a standard benchmark, and has been accepted by CVIU (Paper).

Decomposition of Multiple Foreground Co-segmentation

We formulated our objective function at the superpixel level rather than the pixel level as the traditional optical flow method did.

Our method achieves better large displacement matching capability than LDOF in videos with lower quality , and has been accepted to ICIP 2013 (Paper, Poster).

Superpixel-Based Large Displacement Optical Flow

We applied dictionary learning, visual and motion saliency to extract the foreground object from videos. This research has been accepted to TIP (Paper).

Video Object Extraction Using Saliency

We extract answering logs of the exercises from Junyi Academy (http://www.junyiacademy.org/), an E-learning website similar to Khan Academy.

We use crowdsourcing and machine learning to discover relationships between exercises. Based on that, we will design a mechanism of adaptive test to improve learning experiences of Junyi academy. (Paper, PresentationDemo, Dataset)

Student Modeling and Prerequisite Verification in Knowledge Tree


Computer Vision and Multimedia 

In a single surveillance video, we incorporate tracking data into the photo pop-up algorithm, and utilize RANSAC (random sample consensus) to reduce around 80% of tracking noise (PPT & Videos).

Reasoning 3-D Information from 2-D Images by Tracking

Course Projects 

Active Sampling for estimating QoE model

We use Bayesian learning to model the non-linear relationships between quality of experience (QoE) and multiple factors.

Our experiment shows that active sampling can be used to reduce the number of samples collected from crowdsourcing for building such model.


TAC-KBP is one of the most challenging text-based information retrieval tasks. We integrate research which is done in UMASS IESL in the past year, including embedding linker, multilingual Universal Schema, and LSTM sentence embedding. We perform extensive error analysis and develop some novel techniques (such as using a search engine to reduce noise in training data) to tackle the problems (Paper).

UMASS TAC 2016 system for relation extraction

Neural Network

Inspired by active learning, we propose two alternatives to re-weight training samples based on lightweight estimates of sample uncertainty in stochastic gradient descent (SGD). Extensive experimental results on six datasets show that our methods reliably improve accuracy in various network architectures, including additional gains on top of other popular training techniques (Paper).

Use Active Learning to Improve SGD

Nature Language Processing​

Distributional Inclusion Vector Embedding

We propose a novel word embedding method which preserves the distributional inclusion property in the sparse bag of word (SBOW) feature. The embedding can be used to predict generality of words, detect the hypernym relation, and discover the topics from the raw text simultaneously. The extensive experiments show that the embedding effectively compresses the SBOW, and achieve new state-of-the-art performance on unsupervised hypernym detection tasks (Paper).

(Wix-Logo) Create a site!
100s of templates
No coding needed
Start now >>
(Wix Logo)
This site was created using
Create your own for FREE >>