Jie Song

Jie Song

Auto-generated profile, based on publicly available information

Jie Song

Zhejiang University

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

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

Representative Publications

View All Publications
3 May 2022 in IEEE Transactions on Image Processing
Verified
Journal Article
Closed access
JSJie Song
Jie Song
YCYing Chen
Ying Chen
JYJingwen Ye
Jingwen Ye
MSMingli Song
Mingli Song
Knowledge distillation (KD) has become a well established paradigm for compressing deep neural networks. The typical way of conducting knowledge distillation is to train the student network under the supervision of the teacher network to harness the knowledge at one or multiple spots ( i.e. , layers) in the teacher network. The distillation spots, once specified, will not change for all the training samples, throughout the whole distillation process. In this work, we argue that distillation spots should be adaptive to training samples and distillation epochs. We thus propose a new distillation strategy, termed spot-adaptive KD (SAKD), to adaptively determine the distillation spots in the teacher network per sample, at every training iteration during the whole distillation period. As SAKD actually focuses on “where to distill” instead of “what to distill” that is widely investigated by most existing works, it can be seamlessly integrated into existing distillation methods to further improve their performance. Extensive experiments with 10 state-of-the-art distillers are conducted to demonstrate the effectiveness of SAKD for improving their distillation performance, under both homogeneous and heterogeneous distillation settings. Code is available at https://github.com/zju-vipa/spot-adaptive-pytorch .

22 October 2022
Verified
Book Chapter
Closed access
PDF file
RMRang Meng
Rang Meng
XLXianfeng Li
Xianfeng Li
WCWeijie Chen
Weijie Chen
SYShicai Yang
Shicai Yang
JSJie Song
Jie Song
SPShiliang Pu
Shiliang Pu

20 September 2022 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
KCKaixuan Chen
Kaixuan Chen
JSJie Song
Jie Song
SLShunyu Liu
Shunyu Liu
NYNa Yu
Na Yu
ZFZunlei Feng
Zunlei Feng
MSMingli Song
Mingli Song
Graph-level representation learning is the pivotal step for downstream tasks that operate on the whole graph. The most common approach to this problem is graph pooling, where node features are typically averaged or summed to obtain the graph representations. However, pooling operations like averaging or summing inevitably cause severe information missing, which may severely downgrade the final performance. In this paper, we argue what is crucial to graph-level downstream tasks includes not only the topological structure but also the distribution from which nodes are sampled. Therefore, powered by existing Graph Neural Networks (GNN), we propose a new plug-and-play pooling module, termed as Distribution Knowledge Embedding (DKEPool), where graphs are viewed as distributions on top of GNNs and the pooling goal is to summarize the entire distribution information instead of retaining a certain feature vector by simple predefined pooling operations. A DKEPool network de facto disassembles representation learning into two stages, structure learning and distribution learning . Structure learning follows a recursive neighborhood aggregation scheme to update node features where structure information is obtained. Distribution learning, on the other hand, omits node interconnections and focuses more on the distribution depicted by all the nodes. Extensive experiments on graph classification tasks demonstrate that the proposed DKEPool significantly and consistently outperforms the state-of-the-art methods. The code is avaliable at https://github.com/chenchkx/dkepool

3 November 2022
Verified
Book Chapter
Closed access
PDF file
JYJingwen Ye
Jingwen Ye
YFYifang Fu
Yifang Fu
JSJie Song
Jie Song
XYXingyi Yang
Xingyi Yang
SLSonghua Liu
Songhua Liu
XWXinchao Wang
Xinchao Wang

6 April 2023 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
TZTongya Zheng
Tongya Zheng
XWXinchao Wang
Xinchao Wang
ZFZunlei Feng
Zunlei Feng
JSJie Song
Jie Song
YHYunzhi Hao
Yunzhi Hao
CCChun Chen
Chun Chen
Temporal graphs exhibit dynamic interactions between nodes over continuous time, whose topologies evolve with time elapsing. The whole temporal neighborhood of nodes reveals the varying preferences of nodes. However, previous works usually generate dynamic representation with limited neighbors for simplicity, which results in both inferior performance and high latency of online inference. Therefore, in this paper, we propose a novel method of temporal graph convolution with the whole neighborhood, namely Temporal Aggregation and Propagation Graph Neural Networks (TAP-GNN). Specifically, we firstly analyze the computational complexity of the dynamic representation problem by unfolding the temporal graph in a message-passing paradigm. The expensive complexity motivates us to design the AP (aggregation and propagation) block, which significantly reduces the repeated computation of historical neighbors. The final TAP-GNN supports online inference in the graph stream scenario, which incorporates the temporal information into node embeddings with a temporal activation function and a projection layer besides several AP blocks. Experimental results on various real-life temporal networks show that our proposed TAP-GNN outperforms existing temporal graph methods by a large margin in terms of both predictive performance and online inference latency. Our code is available at https://github.com/doujiang-zheng/TAP-GNN .

26 June 2023 in Proceedings of the AAAI Conference on Artificial Intelligence
Verified
Journal Article
Open access
PDF file
SLShunyu Liu
Shunyu Liu
YZYihe Zhou
Yihe Zhou
JSJie Song
Jie Song
TZTongya Zheng
Tongya Zheng
KCKaixuan Chen
Kaixuan Chen
MSMingli Song
Mingli Song
Value Decomposition (VD) aims to deduce the contributions of agents for decentralized policies in the presence of only global rewards, and has recently emerged as a powerful credit assignment paradigm for tackling cooperative Multi-Agent Reinforcement Learning (MARL) problems. One of the main challenges in VD is to promote diverse behaviors among agents, while existing methods directly encourage the diversity of learned agent networks with various strategies. However, we argue that these dedicated designs for agent networks are still limited by the indistinguishable VD network, leading to homogeneous agent behaviors and thus downgrading the cooperation capability. In this paper, we propose a novel Contrastive Identity-Aware learning (CIA) method, explicitly boosting the credit-level distinguishability of the VD network to break the bottleneck of multi-agent diversity. Specifically, our approach leverages contrastive learning to maximize the mutual information between the temporal credits and identity representations of different agents, encouraging the full expressiveness of credit assignment and further the emergence of individualities. The algorithm implementation of the proposed CIA module is simple yet effective that can be readily incorporated into various VD architectures. Experiments on the SMAC benchmarks and across different VD backbones demonstrate that the proposed method yields results superior to the state-of-the-art counterparts. Our code is available at https://github.com/liushunyu/CIA.