Yifei Xu

Date of Birth : Sept. 22nd, 1996

E-mail : fei960922@gmail.com

Mailing Address: No. 800 Dongchuan Rd., Shanghai, China

Cell Phone : +86 158 0085 2456

Personal Website : http://fei22.cn

Research Interests
Machine Learning, Computer Vision, Data Mining, Statistical Modeling
Education
2013.09 - Present
  • Shanghai Jiao Tong University, China
  • B. S. Eng. in Computer Science
  • ACM Honor Class, Zhiyuan College (a pilot CS class in China)
  • Zhiyuan College, Shanghai Jiao Tong University, China
  • GPA: 3.8 / 4.0 (Major GPA in first three years) (A+ = 4.3)
2016.07 - 2016.09
  • University of California, Los Angeles, United State
  • Cross-disciplinary Scholars in Science and Technology Program
  • Department of Statistics
  • GPA: 4.0 / 4.0
2013.09 - Present

Undergraduate

B. S. Eng. in Computer Science

ACM Honored Class

ACM Honored Class is a pilot computer science class in China.

Over the past 10 years, ACM students have received hundreds of honors and awards. ACM studnets won the ACM International Student Programming Contest World Championship for three times in 2002, 2005 and 2010.

ACM students has published more than 40 academic papers as the first author in the NIPS, WWW, SIGIR, SIGMOD, SIGKDD, ICML, AAAI and other important international conferences and journals.

Zhiyuan College

Zhiyuan College, within Shanghai Jiao Tong University, is an institude that provides an Elite-education for our students. It aims to train them to become future leaders in science and in technology.

In order to be admitted to Zhiyuan College, a student must be on the top fo more than 17,000 undergraduate students within SJTU. Currently, 461 students are enrolled in Zhiyuan College.

By September 2016, 359 students have graduated from Zhiyuan College, 327 (91%) to pursue further studies, 273 (76%) admitted by world top 100 University listed in QS World University Ranking 2016, and 250 (70%) to pursue Ph.D. degrees.

Shanghai Jiao Tong University

Shanghai Jiao Tong University (SJTU), as one of the higher education institutions which enjoy a long history and a world-renowned reputation in China, is a key university directly under the administration of the Ministry of Education (MOE) of the People's Republic of China and co-constructed by MOE and Shanghai Municipal Government. SJTU has become a comprehensive, research-oriented, and internationalized top university in China.

GPA: 3.80 / 4.0 (Major GPA in first three years) (A+ = 4.3)

Major Course grade A+ / A :
  • Programming
  • Linear Algebra
  • Mathematical Analysis
  • University Physics
  • Science and Technology Innovation
  • Computer Architecture
  • Computer System
  • Course Design on Computer System
  • Cmputing Complexity
  • Machine Learning (Include Statistics)
  • Natural Language Processing
  • Database Systems
  • Lab Practice
2016.07 - 2016.09
  • University of California, Los Angeles, United State
  • Department of Statistics

Cross-disciplinary Scholars in Science and Technology Program

The CSST office administers the CSST Summer Program which brings outstanding third year undergraduate students, interested in PhD studies, nominated by top-tier universities in the People?s Republic of China (PRC) and Japan, to conduct 10 week intensive research training with UCLA faculty mentors. This 10-week program offers emerging scholars premier research training in a cutting edge scientific environment that fosters cross-disciplinary collaborations.

GPA: 4.0 / 4.0

Course grade A+ / A :
  • CSST Project
  • Directed Research
Research Experience

Computer and Machine Intelligence Lab

Shanghai Jiao Tong University

2015.07 - Present
Advisor: Liqing Zhang

Research Assistant

  • Large-scale image retrieval competition
  • Model: A model with saliency detection, image classification and image retrieval
  • Implemented Saliency Detection combining Dense and Sparse Reconstruction by Bayesian Integration
  • Classified large-scale images by SVM and Convolution Neural Network
  • Interactive Image Search for Clothing Recommendation
  • Model: Hybrid Topics Model, An LDA based model integrates both visual and text information
  • Used multi-trained Fast-RCNN to localize regions
  • Extracted 3 types of visual descriptors: HOG, LBP, Color
  • Implemented Hybrid Topics Model and introduced a demand-adaptive retrieval strategy

Center for Vision, Cognition, Learning and Autonomy

University of California, LA

2016.07 - 2016.09
Advisor: Ying Nian Wu

Research Intern

  • Learning Generative ConvNet with Continuous Latent Factors
  • Model: a non-linear generalization of factor analysis where the mapping is parametrized by CNN
  • Optimized image synthesis training on large-scale images by batch normalization
  • Used new Back-Propagation inferenced by gradient descent / Langevin dynamics
  • Generative Hierarchical Structure Learning of Sparse FRAME Models
  • Model: Sparse FRAME, a multi-layer probability distribution model captured the part deformation
  • Designed experiments for Sparse FRAME model on detection and clustering
  • Compared Sparse FRAME model with DPM, And-or Graph on point, part, object level detection

Visual Computing Lab

Microsoft Research in Asia

2016.09 - Present
Advisor: Fang Wen

Research Intern

  • Joint Face Detection and Alignment via Cascaded Compositional Learning
  • Model: Sparse FRAME, a multi-layer probability distribution model captured the part deformation
  • Jointed cascade face detection and alignment by advanced boosting algorithm
  • Considered multi-domain to overcome unconstrained face data
  • Trained multi domain on same random forest with both detection and alignment in parallel

Computer and Machine Intelligence Lab

Shanghai Jiao Tong University

2015.07 - Present
Advisor: Liqing Zhang

Research Assistant

*1

Large-scale image retrieval competition

  • Model: A model with saliency detection, image classification and image retrieval
  • Implemented Saliency Detection combining Dense and Sparse Reconstruction by Bayesian Integration
  • Classified large-scale images by SVM and Convolution Neural Network

This is a competition hold by Alibaba. The goal is to output the picture with the most similarity by the given picture. The database is a million web pictures. There are three part for our model. They are saliency detection, CNN classification and text matching. I am in charge of saliency detection and classification. Our team ranked in the top 16 in the competition(Over 2000 teams).

Interactive Image Search for Clothing Recommendation

  • Model: Hybrid Topics Model, An LDA based model integrates both visual and text information
  • Used multi-trained Fast-RCNN to localize regions
  • Extracted 3 types of visual descriptors: HOG, LBP, Color
  • Implemented Hybrid Topics Model and introduced a demand-adaptive retrieval strategy

Paper Abstract

This paper proposes a novel approach to meet users' multi-dimensional requirements in clothing image retrieval.We propose the Hybrid Topic (HT) model to learn the intricate semantic representation of the descriptors above. The model provides an effective multi-dimensional representation of clothes and is able to perform automatic image annotation by probabilistic reasoning from image search. Furthermore, we develop a demand-adaptive retrieval strategy which refines users' specific requirements and removes users' unwanted features. Our experiments show that the HT method significantly outperforms the deep neural network methods.

Center for Vision, Cognition, Learning and Autonomy

University of California, LA

2016.07 - 2016.09
Advisor: Yingnian Wu

Research Intern

Learning Generative ConvNet with Continuous Latent Factors

  • Model: a non-linear generalization of factor analysis where the mapping is parametrized by CNN
  • Optimized image synthesis training on large-scale images by batch normalization
  • Used new Back-Propagation inferenced by gradient descent / Langevin dynamics

Paper Abstract

This paper proposes an alternating back-propagation algorithm for learning the generator network model. The model is a non-linear generalization of factor analysis. In this model, the mapping from the latent factors to the observed vector is parametrized by a convolutional neural network. The alternating back-propagation algorithm iterates between the following two steps: (1) Inferential back-propagation, which infers the latent factors by Langevin dynamics or gradient descent. (2) Learning back-propagation, which updates the parameters given the inferred latent factors by gradient descent.

  • The project page : Link
  • The paper online : Link *My name is listed in the Acknowledgement
  • The poster : Link
  • The presentation : Link

Generative Hierarchical Structure Learning of Sparse FRAME Models

  • Model: Sparse FRAME, a multi-layer probability distribution model captured the part deformation
  • Designed experiments for Sparse FRAME model on detection and clustering
  • Compared Sparse FRAME model with DPM, And-or Graph on point, part, object level detection

Paper Abstract

This paper proposes a framework for generative learning of hierarchical structure of visual objects, based on training hierarchical random field models. The resulting model, which we call structured sparse FRAME model, is a straightforward variation on decomposing the original sparse FRAME model into multiple parts that are allowed to shift their locations, orientations and scales, so that the resulting model becomes a reconfigurable template.

  • The paper online : Link

Visual Computing Lab

Microsoft Research in Asia

2016.09 - Present
Advisor: Fang Wen

Research Intern

*2

Joint Face Detection and Alignment via Cascaded Compositional Learning

  • Model: Sparse FRAME, a multi-layer probability distribution model captured the part deformation
  • Jointed cascade face detection and alignment by advanced boosting algorithm
  • Considered multi-domain to overcome unconstrained face data
  • Trained multi domain on same random forest with both detection and alignment in parallel

Current Working with This work is based on "Joint cascade face detection and alignment" and "Unconstrained Face Alignment via Cascaded Compositional Learning". We aim to provide domain partition on the Joint cascade face detection and alignment method.

Publication
2016.05

Zhengzhong Zhou, Yifei Xu, Jingjin Zhou and Liqing Zhang "Interactive Image Search for Clothing Recommendation. " Proceedings of the 24th ACM international conference on Multimedia. ACM, 2016.

2016.10

Jianwen Xie, Yifei Xu, Ying Nian Wu "Generative Hierarchical Structure Learning of Sparse FRAME Models" Submitted CVPR. 2017.

Honors and Awards
2014.10

Academic Excellence Scholarship at SJTU Prize B (Top 10% in University)

2015.10

Academic Excellence Scholarship at SJTU Prize C (Top 20% in University)

2016.04

Interdisciplinary Contest In Modeling 2016 Meritorious

2016.07

UCLA CSST Scholarship and CSST Award (2 in CSST Program CS Major)

2016.10

'ele' Scholarship for outstanding CS students (6 in university each year)

2016.11

Academic Excellence Scholarship at SJTU Prize B (Top 10% in University)

2016.12

'YuanKang' Scholarship for outstanding research (5 in university each year)

Project Experience
2013.12
"FishTank" Game AI
2014.3
Bookex System (Part)
2014.6
C++ STL Container
2014.8
"Texas Hold'em" Game AI
2014.8
ACM New Website
2015.4
Compiler for simplified C
2015.6
Simulated Pipeline CPU
2015.8
Trajectory Compression
2015.10
Virus for Linux
2016.05
POI System
2016.05
Implicit Discourse Parsing
2016.06
SQL System
2013.12
"FishTank" Game AI
Lang: C++

Project of "Programming"

2014.3
Bookex System (Part)
Lang: Html + PHP + JS

A recommended system for a secondhand book market

2014.6
C++ STL Container
Lang: C++

Project of "Data Struct" which include AVL tree, Hashmap, Linklist, etc.

2014.8
"Texas Hold'em" Game AI
Lang: C++

Project of "Programming Practice"

2014.8
ACM New Website
Lang: Html + PHP + JS

A new, Responsive website for ACM Class.

2015.4
Compiler for simplified C
Lang: Java

A compiler which transform C code into MIPS code.

2015.6
Simulated Pipeline CPU
Lang: Verilog

Simulate the MIPS code’s running on Verilog simulator.

2015.8
Trajectory Compression
Lang: C++

Compress a trajectory with lossless and lossy method.

2015.10
Virus for Linux
Lang: Linux C

A virus runs on Linux in order to have the super authority.

2016.05
POI System
Lang: Html + JSP + JS

A yelp-like website.

2016.05
Implicit Discourse Parsing
Lang: Python + Matlab

The implementation of "Recognizing Implicit Discourse Relations in the Penn Discourse Treebank".

2016.06
SQL System
Lang: C++

A SQL System.

Extracurricular Experience
2013-2014
Member of Zhiyuan Pandeng (leadership) Project
2014-2015
Minisiter of the college publicity center
2014-2017
Publicity commissary and Vice monitor for ACM 2013 Class
2015-2016
Teach Assistant for "Data Structure" (Lead TA)
Qualifications
2014.06

CET-4 611 (Reading 204; Listening 223; Writing 184)

2014.12

CET-6 490 (Reading 187; Listening 192; Writing 111)

2016.05

GRE 324 + 3.5 (Verbal 154; Quantity 170; Writing 3.5)

2016.09
TOEFL 100 /120 (Reading 26; Listening 26; Speaking 23; Writing 25)

*Highest combined score from two separate exams: 96(25,26,20,25); 97(26,24,23,24)