Weifeng Lv

Weifeng Lv

Auto-generated profile, based on publicly available information

Weifeng Lv

Beihang University

Research Focus & Metrics

Academic Career Length

21Years (2006-Present)

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

Representative Publications

View All Publications
19 May 2020 in IEEE Transactions on Image Processing
Verified
Journal Article
Open access
Yang ZhangYang Zhang
Yang Zhang
HSHao Sheng
Hao Sheng
YWYubin Wu
Yubin Wu
SWShuai Wang
Shuai Wang
WLWeifeng Lyu
Weifeng Lyu
ZXZhang Xiong
Zhang Xiong
Recently, most multiple object tracking (MOT) algorithms adopt the idea of tracking-by-detection. Relevant research shows that the performance of the detector obviously affects the tracker, while the improvement of detector is gradually slowing down in recent years. Therefore, trackers using tracklet (short trajectory) are proposed to generate more complete trajectories. Although there are various tracklet generation algorithms, the fragmentation problem still often occurs in crowded scenes. In this paper, we introduce an iterative clustering method that generates more tracklets while maintaining high confidence. Our method shows robust performance on avoiding internal identity switch. Then we propose a deep association method for tracklet association. In terms of motion and appearance, we construct motion evaluation network (MEN) and appearance evaluation network (AEN) to learn long-term features of tracklets for association. In order to explore more robust features of tracklets, a tracklet-based training mechanism is also introduced. Tracklet groups are used as the input of the networks instead of discrete detections. Experimental results show that our training method enhances the performance of the networks. In addition, our tracking framework generates more complete trajectories while maintaining the unique identity of each target as the same time. On the latest MOT 2017 benchmark, we achieve state-of-the-art results.

24 January 2020 in IEEE Transactions on Intelligent Transportation Systems
Verified
Journal Article
Closed access
BDBowen Du
Bowen Du
XHXiao Hu
Xiao Hu
LSLeilei Sun
Leilei Sun
JLJunming Liu
Junming Liu
YQYanan Qiao
Yanan Qiao
WLWeifeng Lv
Weifeng Lv

24 January 2018 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
YTYongxin Tong
Yongxin Tong
LCLei Chen
Lei Chen
ZZZimu Zhou
Zimu Zhou
HJHosagrahar Visvesvaraya Jagadish
Hosagrahar Visvesvaraya Jagadish
LSLidan Shou
Lidan Shou
WLWeifeng Lv
Weifeng Lv
Crowdsourcing has been shown to be effective in a wide range of applications, and is seeing increasing use. A large-scale crowdsourcing task often consists of thousands or millions of atomic tasks, each of which is usually a simple task such as binary choice or simple voting. To distribute a large-scale crowdsourcing task to limited crowd workers, a common practice is to pack a set of atomic tasks into a task bin and send to a crowd worker in a batch. It is challenging to decompose a large-scale crowdsourcing task and execute batches of atomic tasks, which ensures reliable answers at a minimal total cost. Large batches lead to unreliable answers of atomic tasks, while small batches incur unnecessary cost. In this paper, we investigate a general crowdsourcing task decomposition problem, called the Smart Large-scAle task DEcomposer (SLADE) problem, which aims to decompose a large-scale crowdsourcing task to achieve the desired reliability at a minimal cost. We prove the NP-hardness of the SLADE problem and propose solutions in both homogeneous and heterogeneous scenarios. For the homogeneous SLADE problem, where all the atomic tasks share the same reliability requirement, we propose a greedy heuristic algorithm and an efficient and effective approximation framework using an optimal priority queue (OPQ) structure with provable approximation ratio. For the heterogeneous SLADE problem, where the atomic tasks can have different reliability requirements, we extend the OPQ-based framework leveraging a partition strategy, and also prove its approximation guarantee. Finally, we verify the effectiveness and efficiency of the proposed solutions through extensive experiments on representative crowdsourcing platforms.

13 March 2020 in IEEE Internet of Things Journal
Verified
Journal Article
Open access
HSHao Sheng
Hao Sheng
YZYanwei Zheng
Yanwei Zheng
WKWei Ke
Wei Ke
Dongxiao YuDongxiao Yu
Dongxiao Yu
XCXiuzhen Cheng
Xiuzhen Cheng
ZXZhang Xiong
Zhang Xiong
Person re-identification (ReID) is an important application of Internet of Things (IoT). ReID recognizes pedestrians across camera views at different locations and time, which is usually treated as a ranking task. An essential part of this task is the hard sample mining. Technically, two strategies could be employed, i.e., global hard mining and local hard mining. For the former, hard samples are mined within the entire training set, while for the latter, it is done in mini-batches. In literature, most existing methods operate locally. Examples include batch-hard sample mining and semi-hard sample mining. The reason for the rare use of global hard mining is the high computational complexity. In this paper, we argue that global mining helps to find harder samples that benefit model training. To this end, this paper introduces a new system to 1) efficiently mine hard samples (positive and negative) from the entire training set, and 2) effectively use them in training. Specifically, a ranking list network coupled with a multiplet loss is proposed. On one hand, the multiplet loss makes the ranking list progressively created to avoid the time-consuming initialization. On the other hand, the multiplet loss aims to make effective use of the hard and easy samples during training. In addition, the ranking list makes it possible to globally and effectively mine hard positive and negative samples. In the experiments, we explore the performance of the global and local sample mining methods, and the effects of the semi-hard, the hardest and the randomly selected samples. Finally, we demonstrate the validity of our theories using various public datasets and achieve competitive results via a quantitative evaluation.

8 July 2024 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
HHHan Huang
Han Huang
LSLeilei Sun
Leilei Sun
BDBowen Du
Bowen Du
WLWeifeng Lv
Weifeng Lv
Designing new molecules is essential for drug discovery and material science. Recently, deep generative models that aim to model molecule distribution have made promising progress in narrowing down the chemical research space and generating high-fidelity molecules. However, current generative models only focus on modeling 2-D bonding graphs or 3-D geometries, which are two complementary descriptors for molecules. The lack of ability to jointly model them limits the improvement of generation quality and further downstream applications. In this article, we propose a joint 2-D and 3-D graph diffusion model (JODO) that generates geometric graphs representing complete molecules with atom types, formal charges, bond information, and 3-D coordinates. To capture the correlation between 2-D molecular graphs and 3-D geometries in the diffusion process, we develop a diffusion graph transformer (DGT) to parameterize the data prediction model that recovers the original data from noisy data. The DGT uses a relational attention mechanism that enhances the interaction between node and edge representations. This mechanism operates concurrently with the propagation and update of scalar attributes and geometric vectors. Our model can also be extended for inverse molecular design targeting single or multiple quantum properties. In our comprehensive evaluation pipeline for unconditional joint generation, the experimental results show that JODO remarkably outperforms the baselines on the QM9 and GEOM-Drugs datasets. Furthermore, our model excels in few-step fast sampling, as well as in inverse molecule design and molecular graph generation. Our code is provided in https://github.com/GRAPH-0/JODO.

10 November 2021 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
YTYongxin Tong
Yongxin Tong
DSDingyuan Shi
Dingyuan Shi
YXYi Xu
Yi Xu
WLWeifeng Lv
Weifeng Lv
ZQZhiwei Qin
Zhiwei Qin
XTXiaocheng Tang
Xiaocheng Tang
Ride hailing has become prevailing. Central in ride hailing platforms is taxi order dispatching which involves recommending a suitable driver for each order. Previous works use pure combinatorial optimization solutions for taxi dispatching, which suffer in practice due to complex dynamics of demand and supply and temporal dependency among dispatching decisions. Recent studies try to adopt data-driven method into combinatorial optimization hoping knowledge from history data would help overcome these challenges. Among these attempts, adoption of reinforcement learning shows great promise but current adoptions are a unidirectional integration which restricts the potential performance gains. In this work, we propose \underline{L}earning \underline{T}o \underline{D}ispatch(\algoname), a systematic solution that allows synergic integration of reinforcement learning and combinatorial optimization for large-scale taxi order dispatching. We demonstrate the necessity of online learning and taxi scheduling for reinforcement learning to work in synergy with combinatorial optimization, and devise corresponding algorithms. We also devise many tricks for more efficient calculation of the bipartite matching. Experiments show our methods can improve 36.4% and 42.0% on utility and efficiency at most, respectively. Especially, it achieves state-of-the-art performance in terms of utility.