Ying Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Ying Li has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Jun Wu
Fudan University
4 shared publications - Bin Tan
Jinggangshan University
3 shared publications - Die Hu
Fudan University
2 shared publications Baoye Zhang
College of Electronics and Information Engineering, and also with Key Lab of Ministry of Education in Embedded System and Service Computing, Tongji University
1 shared publication- Chen Guo1 shared publication
Chenmou Wu
1 shared publication
Latest Publications
- 1 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. PP (15209210) , 1-12
Empowering Non-Parametric Network With Intrinsic Attention and Optimal Transport for Few-Shot 3D Segmentation
Current approaches to few-shot 3D point cloud segmentation typically depend on heavily pre-trained and parameterized networks. These methods tend to introduce substantial computational costs and exhibit limited generalization ability when applied to new categories across different domains. To address these challenges, we propose a novel Non-Parametric network for few-shot 3D scene Segmentation, NP-3DS. Unlike existing methods that use fixed fusion schemes, our approach proposes a non-parametric intrinsic attention aggregation mechanism that adaptively integrates spatial and chromatic features, effectively preserving fine-grained structures in complex scenes. To mitigate prototype bias and improve cross-domain transferability, we propose an optimal transport prototype calibration strategy that dynamically aligns category representations between support and query sets. Furthermore, to restore structural fidelity during feature upsampling, we design a volume-inverse weighted trilinear interpolation method that captures local geometric correlations, enabling accurate recovery of sharp edges and fine details. Experiments on two benchmarks (S3DIS and ScanNet) demonstrate that NP-3DS significantly outperforms the state-of-the-art non-parametric method (gains +7.37% and +4.48% mIoU, respectively) and matches or surpasses parametric baselines. These results highlight the efficiency and generalizability of our method in a few-shot 3D scene segmentation.
- 1 June 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. PP (15209210) , 1-12
Empowering Non-Parametric Network With Intrinsic Attention and Optimal Transport for Few-Shot 3D Segmentation
Current approaches to few-shot 3D point cloud segmentation typically depend on heavily pre-trained and parameterized networks. These methods tend to introduce substantial computational costs and exhibit limited generalization ability when applied to new categories across different domains. To address these challenges, we propose a novel Non-Parametric network for few-shot 3D scene Segmentation, NP-3DS. Unlike existing methods that use fixed fusion schemes, our approach proposes a non-parametric intrinsic attention aggregation mechanism that adaptively integrates spatial and chromatic features, effectively preserving fine-grained structures in complex scenes. To mitigate prototype bias and improve cross-domain transferability, we propose an optimal transport prototype calibration strategy that dynamically aligns category representations between support and query sets. Furthermore, to restore structural fidelity during feature upsampling, we design a volume-inverse weighted trilinear interpolation method that captures local geometric correlations, enabling accurate recovery of sharp edges and fine details. Experiments on two benchmarks (S3DIS and ScanNet) demonstrate that NP-3DS significantly outperforms the state-of-the-art non-parametric method (gains +7.37% and +4.48% mIoU, respectively) and matches or surpasses parametric baselines. These results highlight the efficiency and generalizability of our method in a few-shot 3D scene segmentation.
- 6 March 2026
- Published by Elsevier in Neurocomputing
- Vol. 679
MIF-gaus: Monocular implicit feature-driven generalizable Gaussian splatting reconstruction
- 6 March 2026
- Published by Elsevier in Neurocomputing
- Vol. 679
MIF-gaus: Monocular implicit feature-driven generalizable Gaussian splatting reconstruction
- Open Access
- 22 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Machine Learning in Communications and Networking
- Vol. 1, 31-42
A Neural Network Empowered Belief Propagation Algorithm Optimized for Short-Cycles in Tanner Graph
Short-cycles in Tanner graphs have a direct impact on the accuracy and effectiveness of the belief propagation (BP) algorithm, as they diverge the BP algorithm by disrupting the independency of message transmission. In this paper, we present a Neural Network Empowered BP (NNE-BP) algorithm as an alternative to the conventional BP algorithm, which can approach the optimum form of BP in shortcycles. The earlier deep learning (DL) based decoder treated the entire decoding process as a classification issue, making it effective only in short block lengths and with limited generalization. In the proposed NNE-BP, we design two neural networks, namely the horizontal network and vertical network, to learn, respectively, the optimized horizontal and vertical processes for single nodes in the presence of cycles. These two networks operate in an iterative fashion, allowing them to approach a theoretical message propagation equation without assuming independent message transmission. Compared to the classification based DL decoder, the proposed network topology is no longer bundled with any specific Tanner graph, and the number of network parameters are greatly reduced. The obtained experimental results show that the proposed NNE-BP outperforms the conventional BP decoder, especially in the case of short-cycles.
- 22 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Machine Learning in Communications and Networking
- Vol. 1, 31-42
Open AccessA Neural Network Empowered Belief Propagation Algorithm Optimized for Short-Cycles in Tanner Graph
Short-cycles in Tanner graphs have a direct impact on the accuracy and effectiveness of the belief propagation (BP) algorithm, as they diverge the BP algorithm by disrupting the independency of message transmission. In this paper, we present a Neural Network Empowered BP (NNE-BP) algorithm as an alternative to the conventional BP algorithm, which can approach the optimum form of BP in shortcycles. The earlier deep learning (DL) based decoder treated the entire decoding process as a classification issue, making it effective only in short block lengths and with limited generalization. In the proposed NNE-BP, we design two neural networks, namely the horizontal network and vertical network, to learn, respectively, the optimized horizontal and vertical processes for single nodes in the presence of cycles. These two networks operate in an iterative fashion, allowing them to approach a theoretical message propagation equation without assuming independent message transmission. Compared to the classification based DL decoder, the proposed network topology is no longer bundled with any specific Tanner graph, and the number of network parameters are greatly reduced. The obtained experimental results show that the proposed NNE-BP outperforms the conventional BP decoder, especially in the case of short-cycles.
- 14 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 68 (12) , 11944-11954
Compressive Spinal Codes
It is well known that spinal codes can essentially approach channel capacity for equiprobable source, and achieve seamless rate adaptation at the same time. But how to design an optimal spinal codes for non-equiprobable source is still an open problem. This paper proposes a novel Sparse Spinal (S-Spinal) codes design, which can essentially approach channel capacity for non-equiprobable and compressible source, and its seamless rate adaptation is also preserved. The main idea is to utilize source distribution information at the spinal decoder. The sparser the source is, the less symbols the receiver needs. The rateless feature of spinal codes can provide fine-grain compression ratio. The different decoder designs for Binary Symmetrical Channel (BSC) and Additive White Gaussian Noise (AWGN) channel are proposed, since their priori information is different. We also analyze and prove the S-Spinal codes with a practical pruning decoder can essentially achieve the channel capacity. The experimental results show that the proposed S-Spinal is very close to channel capacity of both BSC and AWGN channel for non-equiprobable source, and can achieve seamless rate adaption in practical system.
- 14 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 68 (12) , 11944-11954
Compressive Spinal Codes
It is well known that spinal codes can essentially approach channel capacity for equiprobable source, and achieve seamless rate adaptation at the same time. But how to design an optimal spinal codes for non-equiprobable source is still an open problem. This paper proposes a novel Sparse Spinal (S-Spinal) codes design, which can essentially approach channel capacity for non-equiprobable and compressible source, and its seamless rate adaptation is also preserved. The main idea is to utilize source distribution information at the spinal decoder. The sparser the source is, the less symbols the receiver needs. The rateless feature of spinal codes can provide fine-grain compression ratio. The different decoder designs for Binary Symmetrical Channel (BSC) and Additive White Gaussian Noise (AWGN) channel are proposed, since their priori information is different. We also analyze and prove the S-Spinal codes with a practical pruning decoder can essentially achieve the channel capacity. The experimental results show that the proposed S-Spinal is very close to channel capacity of both BSC and AWGN channel for non-equiprobable source, and can achieve seamless rate adaption in practical system.
- 24 May 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 7 (6) , 930-933
A Hybrid Digital-Analog Chaotic Code
Traditional analog chaotic codes suffer from the vulnerability of signs at the receiver, and this situation deteriorates at low signal-to-noise ratio (SNR). In order to solve this problem, this letter proposes a hybrid digital-analog chaotic code (HDACC), which uses chaotic states (analog value) and part of their signs (digital value) together as parity codewords. These signs are further encoded in a digital way. We derived the theoretical decoding Mean Square Error (MSE) of HDACC over additive white Gaussian noise (AWGN) channel and gave the optimal digital/analog codewords allocation. Simulation results demonstrate that the derived theoretical MSE matches very well with simulated MSE curves. Evaluations show that HDACC outperforms existing related works in a wide SNR range.
- 24 May 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 7 (6) , 930-933
A Hybrid Digital-Analog Chaotic Code
Traditional analog chaotic codes suffer from the vulnerability of signs at the receiver, and this situation deteriorates at low signal-to-noise ratio (SNR). In order to solve this problem, this letter proposes a hybrid digital-analog chaotic code (HDACC), which uses chaotic states (analog value) and part of their signs (digital value) together as parity codewords. These signs are further encoded in a digital way. We derived the theoretical decoding Mean Square Error (MSE) of HDACC over additive white Gaussian noise (AWGN) channel and gave the optimal digital/analog codewords allocation. Simulation results demonstrate that the derived theoretical MSE matches very well with simulated MSE curves. Evaluations show that HDACC outperforms existing related works in a wide SNR range.