Qizhen Weng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Qizhen Weng has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Jiaocheng Hu
School of Electronics and Communication Engineering, Sun Yat-sen University, Shenzhen, China
2 shared publications - Xiangwei Zhu
Sun Yat-sen University
2 shared publications Can Gong
1 shared publication- Dingwei Peng1 shared publication
Gelu Liu
1 shared publication- Gelu Liu
School of Electronics and Communication Engineering, Sun Yat-sen University, Shenzhen, China
1 shared publication
Latest Publications
- 10 November 2026
- Published by Elsevier in Signal Processing
- Vol. 253
KNet-Suite: Improving efficiency and robustness in neural network-enhanced Kalman filtering under unknown noise statistics
- 10 November 2026
- Published by Elsevier in Signal Processing
- Vol. 253
KNet-Suite: Improving efficiency and robustness in neural network-enhanced Kalman filtering under unknown noise statistics
- 17 December 2024
- Published by Elsevier in Computers and Electrical Engineering
- Vol. 122
LMCodec2: Ultra-low bit rate codec with causal multiple transformers
- 17 December 2024
- Published by Elsevier in Computers and Electrical Engineering
- Vol. 122
LMCodec2: Ultra-low bit rate codec with causal multiple transformers
- 5 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 9 (11) , 9510-9517
A New Representation of Universal Successor Features for Enhancing the Generalization of Target-Driven Visual Navigation
Target-driven visual navigation is a long-standing objective in the field of robotics. Deep reinforcement learning methods have demonstrated their effectiveness in developing target-driven visual navigation policies, yet they often struggle with generalization. Although extended reinforcement learning methods based on successor features enhance cross-target generalization, they falter in unseen environments. To tackle this issue, we put forth an approach that builds upon successor features and introduces two novel designs: the parsimonious dynamics and the adaptive policy. Specifically, we employ state features that predict successor features to create latent parsimonious dynamics models. Furthermore, learning a set of rules or dynamics within the latent state facilitates the identification of informative features for prediction and the acquisition of their corresponding successor features. The adaptive policy is to prevent the agent getting trapped by repeatedly taking the same action repeatedly even though it doesn't lead to any progress. Comprehensive results from the interactive AI2-THOR and the Habitat simulator show that agents trained using this method achieve state-of-the-art capabilities in generalization across target, scene, and domain. Additionally, we propose a continuous learning experiment that utilizes pre-trained models to achieve superior generalization.
- 5 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 9 (11) , 9510-9517
A New Representation of Universal Successor Features for Enhancing the Generalization of Target-Driven Visual Navigation
Target-driven visual navigation is a long-standing objective in the field of robotics. Deep reinforcement learning methods have demonstrated their effectiveness in developing target-driven visual navigation policies, yet they often struggle with generalization. Although extended reinforcement learning methods based on successor features enhance cross-target generalization, they falter in unseen environments. To tackle this issue, we put forth an approach that builds upon successor features and introduces two novel designs: the parsimonious dynamics and the adaptive policy. Specifically, we employ state features that predict successor features to create latent parsimonious dynamics models. Furthermore, learning a set of rules or dynamics within the latent state facilitates the identification of informative features for prediction and the acquisition of their corresponding successor features. The adaptive policy is to prevent the agent getting trapped by repeatedly taking the same action repeatedly even though it doesn't lead to any progress. Comprehensive results from the interactive AI2-THOR and the Habitat simulator show that agents trained using this method achieve state-of-the-art capabilities in generalization across target, scene, and domain. Additionally, we propose a continuous learning experiment that utilizes pre-trained models to achieve superior generalization.