Fanyu Zeng

Fanyu Zeng

Auto-generated profile, based on publicly available information

Fanyu Zeng

Nanjing University of Posts and Telecommunications

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
23 July 2020 in IEEE Access
Verified
Journal Article
Open access
FZFanyu Zeng
Fanyu Zeng
CWChen Wang
Chen Wang
SGShuzhi Sam Ge
Shuzhi Sam Ge
Visual navigation (vNavigation) is a key and fundamental technology for artificial agents’ interaction with the environment to achieve advanced behaviors. Visual navigation for artificial agents with deep reinforcement learning (DRL) is a new research hotspot in artificial intelligence and robotics that incorporates the decision making of DRL into visual navigation. Visual navigation via DRL, an end-to-end method, directly receives the high-dimensional images and generates an optimal navigation policy. In this paper, we first present an overview on reinforcement learning (RL), deep learning (DL) and deep reinforcement learning (DRL). Then, we systematically describe five main categories of visual DRL navigation: direct DRL vNavigation, hierarchical DRL vNavigation, multi-task DRL vNavigation, memory-inference DRL vNavigation and vision-language DRL vNavigation. These visual DRL navigation algorithms are reviewed in detail. Finally, we discuss the challenges and some possible opportunities to visual DRL navigation for artificial agents.

11 December 2023 in IEEE Transactions on Computational Imaging
Verified
Journal Article
Closed access
GHGuang Han
Guang Han
KWKang Wu
Kang Wu
FZFanyu Zeng
Fanyu Zeng
JLJixin Liu
Jixin Liu
KWONG Tak Wu SamKWONG Tak Wu Sam
KWONG Tak Wu Sam
Low-light image enhancement is a crucial area of research in computer vision, aimed at recovering normally exposed images from low-light images to facilitate high-level vision tasks such as target detection, tracking, and recognition. However, the convolutional neural networks commonly used in this field have a bias towards extracting low-frequency local structural features in the spatial domain, resulting in unclear texture details in the enhanced images. To address this limitation, this paper proposes a novel frequency-domain perception-based model, called DSFPNet. This model has three unique features. First, the recursive The Hadamard product method is used to model long-range relationships of the Transformer in the feature extraction. Second, a bilateral gating mechanism is added to the feedforward network to filter out useless information and improve the nonlinear modeling capability of the module. Third, cross-layer connectivity of low-frequency features is added to maintain the structural stability of features at different levels. In addition, the proposed model uses an EFF (Enhancing Frequency Features) module to extract frequency features and selectively fuse the three high frequencies. The fused features are then aligned and reconstructed with the low-frequency features to obtain the final frequency features, which helps the model recover enhanced image texture details. Through a large number of experimental results on the two public datasets, the proposed DSFPNet model shows excellent performances and outperforms many existing state-of-the-art methods. The model exhibits good potential and is expected to advance the field of low-light image enhancement in computer vision.

14 October 2020 in Journal of Robotics
Verified
Journal Article
Open access
PDF file
FZFanyu Zeng
Fanyu Zeng
CWChen Wang
Chen Wang
Vanilla policy gradient methods suffer from high variance, leading to unstable policies during training, where the policys performance fluctuates drastically between iterations. To address this issue, we analyze the policy optimization process of the navigation method based on deep reinforcement learning (DRL) that uses asynchronous gradient descent for optimization. A variant navigation (asynchronous proximal policy optimization navigation, appoNav) is presented that can guarantee the policy monotonic improvement during the process of policy optimization. Our experiments are tested in DeepMind Lab, and the experimental results show that the artificial agents with appoNav perform better than the compared algorithm.

24 November 2020 in IEEE Transactions on Cognitive and Developmental Systems
Verified
Journal Article
Closed access
FZFanyu Zeng
Fanyu Zeng
CWChen Wang
Chen Wang
SGShuzhi Ge
Shuzhi Ge
Traditional reinforcement learning makes policy based on the current system state. However, insufficient system information and few rewards lead to its limited applicability, especially in a partially observed environment with sparse rewards. In this work, we propose a tutor-student network (TSN) for improving an agent’s performance with additional auxiliary information. In the tutor-student framework, a tutor module generates auxiliary information, while a student module refers to the tutor’s suggestion during training. The key of our proposed approach is that tutor provides prior knowledge that does not correspond to a specified environment to help the student module accelerate the learning procedure. We build twelve indoor mazes in ViZDoom including empty mazes and mazes with obstacles, evaluate the performance of TSN compared with advantage actor critic (A2C) and show that the proposed network learnt navigation faster and obtained higher accumulated rewards. More importantly, our approach could generalize well to new and unseen domains.

23 March 2023 in Multimedia Tools and Applications
Verified
Journal Article
Closed access
PDF file
GHGuang Han
Guang Han
YZYu Zhou
Yu Zhou
FZFanyu Zeng
Fanyu Zeng

21 September 2020 in Journal of Robotics
Verified
Journal Article
Open access
PDF file
FZFanyu Zeng
Fanyu Zeng
XCXi Cai
Xi Cai
SGShuzhi Sam Ge
Shuzhi Sam Ge
Wall defect detection is an important function for autonomous decoration robots. Object detection methods based on deep neural networks require a large number of images with the handcrafted bounding box for training. Nonetheless, building large datasets manually is impractical, which is time-consuming and labor-intensive. In this work, we solve this issue to propose the low-shot wall defect detection algorithm using deep reinforcement learning (DRL) for autonomous decoration robots. Our algorithm first utilizes the attention proposal network (APN) to generate attention regions and applies AlexNet to extract the features of attention patches to further reduce computation. Finally, we train our method with deep reinforcement learning to learn the optimal detection policy. The experiments are implemented on a low-shot dataset in which images are collected from real decoration environments, and the experimental results show the proposed method can achieve fast convergence and learn the optimal detection policy for wall defect images.