Search Results

Overfitting in RL has become one of the main obstacles to applications in reinforcement learning (RL). Existing methods do not provide explicit semantic constrain for the feature extractor, hindering the agent from learning a unified cross-domain ...

Model-based offline reinforcement learning (RL) algorithms have emerged as a promising paradigm for offline RL. These algorithms usually learn a dynamics model from a static dataset of transitions, use the model to generate synthetic trajectories, and ...

Offline meta-reinforcement learning (OMRL) utilizes pre-collected offline datasets to enhance the agent's generalization ability on unseen tasks. However, the context shift problem arises due to the distribution discrepancy between the contexts used for ...

In the field of multi-task reinforcement learning, the modular principle, which involves specializing functionalities into different modules and combining them appropriately, has been widely adopted as a promising approach to prevent the negative ...

Domain adaptation in reinforcement learning (RL) mainly deals with the changes of observation when transferring the policy to a new environment. Many traditional approaches of domain adaptation in RL manage to learn a mapping function between the source ...

The durability of commercial vehicles is heavily dependent on the design of their suspension systems’ hard points. The design optimization of hard points involves a large number of design parameters. However, traditional surrogate model-assisted ...

Many unavoidable uncertainties in the engineering structure will affect its performance response. It is necessary to analyze the uncertain responses generated by uncertain factors in the vehicle design. Therefore, this study proposes a feasible ...

This study proposes a non-deterministic robust topology optimization of ply orientation for multiple fiber-reinforced plastic (FRP) materials, such as carbon fiber---reinforced plastic (CFRP) and glass fiber---reinforced plastic (GFRP) composites, under ...

Multi-cell thin-walled structures have proven fairly effective in energy absorption and have been extensively used in vehicle engineering. However, the effects of multi-cell configurations and oblique loads on the crashworthiness performance have been ...

Structural optimization for vehicle door signifies one of the key topics of research to continuously improve its performances. However, majority of the studies to date have not considered uncertainties whilst it has been known that a deterministic ...

Matrix embedding is proved an effective way to improve embedding efficiency of steganography, and usually, higher dimensional matrix will provide better embedding efficiency. However, the sender might suffer huge computational complexity when employing ...

Embedding efficiency is a metric of steganographic security. Higher embedding efficiency is usually desired to get more secure steganography. Matrix embedding is proved an effective way to increase embedding efficiency of steganography, and meanwhile ...

In order to enhance the steganographic security, higher embedding efficiency (average number of secret data bits embedded per one embedding change) is desired. Matrix embedding is a well-known steganographic scheme that can improve the embedding ...

In this paper, we present an improved method for detecting LSB matching steganography in gray-scale image. Our improvements focus on three aspects: (1) instead of using the amplitude of local extrema of the image's histogram in the previous work, we turn ...