Haifeng Liu

Haifeng Liu

Auto-generated profile, based on publicly available information

Haifeng Liu

College of Computer Science, Zhejiang University, Hangzhou, China

Research Focus & Metrics

Academic Career Length

19Years (2008-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 November 2011 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
HLHaifeng Liu
Haifeng Liu
ZWZhaohui Wu
Zhaohui Wu
XLXuelong Li
Xuelong Li
DCDeng Cai
Deng Cai
THThomas S. Huang
Thomas S. Huang
Nonnegative matrix factorization (NMF) is a popular technique for finding parts-based, linear representations of nonnegative data. It has been successfully applied in a wide range of applications such as pattern recognition, information retrieval, and computer vision. However, NMF is essentially an unsupervised method and cannot make use of label information. In this paper, we propose a novel semi-supervised matrix decomposition method, called Constrained Nonnegative Matrix Factorization (CNMF), which incorporates the label information as additional constraints. Specifically, we show how explicitly combining label information improves the discriminating power of the resulting matrix decomposition. We explore the proposed CNMF method with two cost function formulations and provide the corresponding update solutions for the optimization problems. Empirical experiments demonstrate the effectiveness of our novel algorithm in comparison to the state-of-the-art approaches through a set of evaluations based on real-world applications.

9 January 2019 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
YZYu Zhu
Yu Zhu
JLJinghao Lin
Jinghao Lin
SHShibi He
Shibi He
BWBeidou Wang
Beidou Wang
Ziyu GuanZiyu Guan
Ziyu Guan
DCDeng Cai
Deng Cai
In recommender systems, cold-start issues are situations where no previous events, e.g. ratings, are known for certain users or items. In this paper, we focus on the item cold-start problem. Both content information (e.g. item attributes) and initial user ratings are valuable for seizing users' preferences on a new item. However, previous methods for the item cold-start problem either (1) incorporate content information into collaborative filtering to perform hybrid recommendation, or (2) actively select users to rate the new item without considering content information and then do collaborative filtering. In this paper, we propose a novel recommendation scheme for the item cold-start problem by leveraging both active learning and items' attribute information. Specifically, we design useful user selection criteria based on items' attributes and users' rating history, and combine the criteria in an optimization framework for selecting users. By exploiting the feedback ratings, users' previous ratings and items' attributes, we then generate accurate rating predictions for the other unselected users. Experimental results on two real-world datasets show the superiority of our proposed method over traditional methods.

23 October 2022
Verified
Book Chapter
Closed access
PDF file
LPLiang Peng
Liang Peng
XWXiaopei Wu
Xiaopei Wu
ZYZheng Yang
Zheng Yang
HLHaifeng Liu
Haifeng Liu
DCDeng Cai
Deng Cai

8 February 2021 in IEEE Transactions on Image Processing
Verified
Journal Article
Closed access
HFHao Feng
Hao Feng
MCMinghao Chen
Minghao Chen
JHJinming Hu
Jinming Hu
DSDong Shen
Dong Shen
HLHaifeng Liu
Haifeng Liu
DCDeng Cai
Deng Cai
In recent years, supervised person re-identification (re-ID) models have received increasing studies. However, these models trained on the source domain always suffer dramatic performance drop when tested on an unseen domain. Existing methods are primary to use pseudo labels to alleviate this problem. One of the most successful approaches predicts neighbors of each unlabeled image and then uses them to train the model. Although the predicted neighbors are credible, they always miss some hard positive samples, which may hinder the model from discovering important discriminative information of the unlabeled domain. In this paper, to complement these low recall neighbor pseudo labels, we propose a joint learning framework to learn better feature embeddings via high precision neighbor pseudo labels and high recall group pseudo labels. The group pseudo labels are generated by transitively merging neighbors of different samples into a group to achieve higher recall. However, the merging operation may cause subgroups in the group due to imperfect neighbor predictions. To utilize these group pseudo labels properly, we propose using a similarity-aggregating loss to mitigate the influence of these subgroups by pulling the input sample towards the most similar embeddings. Extensive experiments on three large-scale datasets demonstrate that our method can achieve state-of-the-art performance under the unsupervised domain adaptation re-ID setting.

23 October 2022
Verified
Book Chapter
Closed access
PDF file
LPLiang Peng
Liang Peng
FLFei Liu
Fei Liu
ZYZhengxu Yu
Zhengxu Yu
SYSenbo Yan
Senbo Yan
DDDan Deng
Dan Deng
DCDeng Cai
Deng Cai

7 April 2014 in International Journal of Parallel, Emergent and Distributed Systems
Verified
Journal Article
Closed access
NANana Yaw Asabere
Nana Yaw Asabere
FXFeng Xia
Feng Xia
QMQinxue Meng
Qinxue Meng
FLFengqi Li
Fengqi Li
HLHaifeng Liu
Haifeng Liu
The significant proliferation of research papers in both conferences and journals has made it difficult for researchers to easily access relevant scholarly papers for academic learning. This has been a substantial problem for many researchers. Conferences, in comparison with journals, have an aspect of social learning and networking, which leads to personal familiarisation through various interactions among researchers. In this paper, we improve the social awareness of conference participants by proposing a novel folksonomy-based paper recommendation algorithm, called socially aware recommendation of scholarly papers (SARSP). SARSP recommends papers issued by active participants (APs), to other Group Profile Participants at the same conference based on preference similarity of their research interests. In addition, SARSP computes the social ties between an AP and other conference participants to effectively generate social recommendations of scholarly papers. We evaluate our proposed algorithm using a real-world data-set. Our experimental results confirm that SARSP has significant improvement over other existing methods.