He Zhang

He Zhang
Aalto University · Department of Computer Science

PhD in Machine Learning

About

23
Publications
20,960
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
131
Citations
Introduction
I have been working as a Data Scientist within industry since 2014. I obtained my PhD in Machine Learning in 2014. My PhD advisor was Professor Erkki Oja, IEEE Life Fellow. I have: *8+ years project and research experience on Machine Learning and Image Processing. *Strong analytical and programming skills using Matlab, Python, SQL, C/C++, Perl, JavaScript. My LinkedIn: fi.linkedin.com/in/ihezhang/
Additional affiliations
January 2008 - present
Aalto University
Position
  • Project Researcher
Description
  • Working on projects related to Machine Learning, Data Mining, Cluster Analysis, Affective Image Processing. I have published over 11 papers and journals, and have given oral presentation in 10 different conferences or research forums.
January 2008 - June 2014
Aalto University
Position
  • Project Researcher on Data Mining, Machine Learning, and Image Processing
July 2007 - September 2007
Aalto University
Position
  • Summer Trainee
Description
  • Working on Audio and Video Synchronizations.
Education
January 2008 - June 2014
Aalto University
Field of study
  • Information and Computer Science
September 2004 - June 2007
Jilin University
Field of study
  • Information and Communication Systems
September 2000 - July 2004
Jilin University
Field of study
  • Communication Engineering

Publications

Publications (23)
Article
Affective classification and retrieval of multimedia such as audio, image, and video have become emerging research areas in recent years. The previous research focused on designing features and developing feature extraction methods. Generally, a multimedia content can be represented with different feature representations (i.e., views). However, the...
Article
Many modern clustering methods employ a non-convex objective function and use iterative optimization algorithms to find local minima. Thus initialization of the algorithms is very important. Conventionally the starting guess of the iterations is randomly chosen; however, such a simple initialization often leads to poor clusterings. Here we propose...
Conference Paper
Full-text available
In this work, we study people’s emotions evoked by viewing abstract art images based on traditional low-level image features within a binary classification framework. Abstract art is used here instead of artistic or photographic images because those contain contextual information that influences the emotional assessment in a highly individual manne...
Conference Paper
Full-text available
We describe a setup and experiments where users are checking and correcting image tags given by an automatic tagging system. We study how much the application of a content-based image retrieval (CBIR) method speeds up the process of finding and correcting the erroneously-tagged images. We also analyze the use of implicit relevance feedback from the...
Article
Affective classification and retrieval of multimedia such as audio, image, and video have become emerging research areas in recent years. The previous research focused on designing features and developing feature extraction methods. Generally, a multimedia content can be represented with different feature representations (i.e., views). However, the...
Data
Full-text available
Conference Paper
Full-text available
Images usually convey information that can influence people’s emotional states. Such affective information can be used by search engines and social networks for better understanding the user’s preferences. We propose here a novel Bayesian multiple kernel learning method for predicting the emotions evoked by images. The proposed method can make use...
Conference Paper
Full-text available
Emotional semantic image retrieval systems aim at incorporating the user’s affective states for responding adequately to the user’s interests. One challenge is to select features specific to image affect detection. Another challenge is to build effective learning models or classifiers to bridge the so-called “affective gap”. In this work, we study...
Conference Paper
Full-text available
Projective Nonnegative Matrix Factorization (PNMF) is able to extract sparse features and provide good approximation for discrete problems such as clustering. However, the original PNMF optimization algorithm can not guarantee theoretical convergence during the iterative learning. We propose here an adaptive multiplicative algorithm for PNMF which...
Conference Paper
Full-text available
Projective Nonnegative Matrix Factorization (PNMF) is one of the recent methods for computing low-rank approximations to data matrices. It is advantageous in many practical application domains such as clustering, graph partitioning, and sparse feature extraction. However, up to now a scalable implementation of PNMF for large-scale machine learning...
Conference Paper
Full-text available
In the past decade, Probabilistic Latent Semantic Indexing (PLSI) has become an important modeling technique, widely used in clustering or graph partitioning analysis. However, the original PLSI is designed for multinomial data and may not handle other data types. To overcome this restriction, we generalize PLSI to t-exponential family based on a r...
Conference Paper
Full-text available
Explicit relevance feedback requires the user to explicitly refine the search queries for content-based image retrieval. This may become laborious or even impossible due to the ever-increasing volume of digital databases. We present a multimodal information collector that can unobtrusively record and asynchronously transmit the user’s implicit rele...
Conference Paper
Full-text available
The I-divergence or unnormalized generalization of Kullback-Leibler (KL) divergence is commonly used in Nonnegative Matrix Factorization (NMF). This divergence has the drawback that its gradients with respect to the factorizing matrices depend heavily on the scales of the matrices, and learning the scales in gradient-descent optimization may requir...
Article
Full-text available
The Proactive Interfaces research theme combines efforts of multiple research groups, including the Statistical Machine Learning and Bioinformatics group, lead by Profes-sor Samuel Kaski, and the Content-Based Information Retrieval and Speech Recognition groups, lead by Professor Erkki Oja. Since 2008, major collaborative EU FP7, EIT ICT Labs and A...
Article
Full-text available
This report presents a literature survey conducted to review the current state of the art in research concerning the use of eye movement measurements and other non-conventional and implicit relevance feedback modalities in content-based image and information retrieval. We define and elaborate on the concept of enriched relevance feedback and study...

Questions

Questions (6)
Question
Any tutorials, papers, and Python libraries related to RNN (especially using RNN for time series analysis, for e.g, clustering, classification, anomaly detections) would be helpful!
Question
Some Harvard guys just sent me an Ads. email to promote their newly-invented scientific writing software - Authorea (www.authorea.com). They claimed its advantages over LaTex, but what do you think, my fellow colleagues?  
Question
Now I have a project that needs me to build the training data and testing data (in this context, data means images) from scratch, so that a classifier (e.g. SVM) can be trained and tested.
Any good criteria to build training and testing images? I do not mean how to extract features or how to split between training and testing images. What I mean is how to collect a good image dataset.  
Question
I would like to extract various image features for phone screenshot images recognition. I hope the feature extraction method runs fast, so perhaps the method should be implemented in Python and / or C ++. Any source code links would be very helpful!
Thanks a lot! 
Question
Images often contain information that can trigger people's affective feelings.

Network

Cited By