Abstract:Integrating low-light image enhancement techniques, in which diffusion-based AI-generated content (AIGC) models are promising, is necessary to enhance nighttime teleoperation. Remarkably, the AIGC model is computation-intensive, thus necessitating the allocation of AIGC tasks to edge servers with ample computational resources. Given the distinct cost of the AIGC model trained with varying-sized datasets and AIGC tasks possessing disparate demand, it is imperative to formulate a differential pricing strategy to optimize the utility of teleoperators and edge servers concurrently. Nonetheless, the pricing strategy formulation is under information asymmetry, i.e., the demand (e.g., the difficulty level of AIGC tasks and their distribution) of AIGC tasks is hidden information to edge servers. Additionally, manually assessing the difficulty level of AIGC tasks is tedious and unnecessary for teleoperators. To this end, we devise a framework of AIGC task allocation assisted by the Vision Language Model (VLM)-empowered contract theory, which includes two components: VLM-empowered difficulty assessment and contract theory-assisted AIGC task allocation. The first component enables automatic and accurate AIGC task difficulty assessment. The second component is capable of formulating the pricing strategy for edge servers under information asymmetry, thereby optimizing the utility of both edge servers and teleoperators. The simulation results demonstrated that our proposed framework can improve the average utility of teleoperators and edge servers by 10.88~12.43% and 1.4~2.17%, respectively. Code and data are available at <a class="link-external link-https" href="https://github.com/ZiJun0819/VLM-Contract-Theory" rel="external noopener nofollow">this https URL</a>.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is how to optimize the assignment of AI - generated content (AIGC) tasks in teleoperation, especially to improve the quality and efficiency of image enhancement under low - light conditions. Specifically, the paper focuses on the following two key issues: 1. **Differentiated Pricing Strategies under Information Asymmetry**: - **Problem Description**: How can the edge server formulate differentiated pricing strategies to optimize the utilities of both the edge server and the remote operator simultaneously without knowing the difficulty distribution of AIGC tasks? - **Challenge**: Before receiving AIGC tasks, the edge server cannot obtain the specific difficulty of the tasks, which leads to the problem of information asymmetry. Therefore, a mechanism is required to ensure that the pricing strategy can not only motivate the edge server to process tasks efficiently but also meet the needs of the remote operator. 2. **Automatically and Accurately Assessing the Difficulty of AIGC Tasks**: - **Problem Description**: How can the visual - language model (VLM) be used to automatically and accurately assess the difficulty of AIGC tasks to ensure the efficiency of the remote operator? - **Challenge**: Manual assessment of task difficulty is both cumbersome and unnecessary, especially in the teleoperation scenario where the operator hopes to operate as easily as during the day without additional effort. Therefore, an automated method is needed to assess task difficulty. ### Solution Overview To solve the above problems, the paper proposes a contract - theory - based framework empowered by the visual - language model (VLM), which consists of two main components: 1. **VLM - Empowered Task Difficulty Assessment**: - Use the trained visual - language model (VLM) to automatically assess the difficulty of AIGC tasks. Through interaction with the VLM agent, determine the optimal assessment indicators and assess the difficulty of different tasks according to these indicators. - For example, the VLM agent can assess the task difficulty based on factors such as the visibility of details in the image, the complexity of shadows, and the overall brightness. 2. **Contract - Theory - Assisted Task Assignment**: - In the case of information asymmetry, use contract theory to formulate differentiated pricing strategies. By defining the contract bundle, i.e., {required AIGC model performance, pricing}, to ensure the maximization of the utilities of the edge server and the remote operator. - Specifically, the edge server will select an appropriate AIGC model according to the task difficulty and formulate the corresponding price. The remote operator will select the most appropriate AIGC service according to the task difficulty. ### Main Contributions - Propose a contract - theory - based framework empowered by VLM for AIGC task assignment, which can optimize the utilities of both the edge server and the remote operator simultaneously under information asymmetry. - Design an automatic and accurate AIGC task difficulty assessment scheme, using the VLM agent for assessment, ensuring the automation and efficiency of task assignment. - Verify the effectiveness of this framework through numerical simulation experiments. The results show that this framework can increase the average utility of the remote operator and the edge server by 10.88% - 12.43% and 1.4% - 2.17% respectively. In conclusion, by combining VLM and contract theory, this paper solves the key problems of AIGC task assignment in teleoperation and improves the quality and efficiency of image enhancement under low - light conditions.

Vision Language Model-Empowered Contract Theory for AIGC Task Allocation in Teleoperation

A Game-Theoretic Approach Based Task Offloading and Resource Pricing Method for Idle Vehicle Devices Assisted VEC

Optimizing AIGC Services by Prompt Engineering and Edge Computing: A Generative Diffusion Model-Based Contract Theory Approach

Joint Device Scheduling and Resource Allocation for ISCC-Based Multi-View-Multi-Task Inference

Task allocation algorithm and optimization model on edge collaboration

Joint Model Assignment and Resource Allocation for Cost-Effective Mobile Generative Services

Efficient Task Allocation for Computation Offloading in Vehicular Edge Computing

Cached Model-as-a-Resource: Provisioning Large Language Model Agents for Edge Intelligence in Space-air-ground Integrated Networks

Efficient Deployment of Large Language Model Across Cloud-Device Systems

Computing Resource Allocation Strategy Based on Cloud-Edge Cluster Collaboration in Internet of Vehicles

Joint task offloading and resource allocation in vehicle-assisted multi-access edge computing

Diffusion Model-based Incentive Mechanism with Prospect Theory for Edge AIGC Services in 6G IoT

BARGAIN-MATCH: A Game Theoretical Approach for Resource Allocation and Task Offloading in Vehicular Edge Computing Networks

Revenue and Energy Efficiency-Driven Delay Constrained Computing Task Offloading and Resource Allocation in a Vehicular Edge Computing Network: A Deep Reinforcement Learning Approach

Secure incentive mechanism for energy trading in computing force networks enabled internet of vehicles: a contract theory approach

Toward Real-Time Edge AI: Model-Agnostic Task-Oriented Communication with Visual Feature Alignment

Simulation-based joint user assignment and edge resource allocation optimization for hybrid tasks in vehicular edge computing

Joint Offloading Decision and Resource Allocation for Vehicular Fog-Edge Computing Networks: A Contract-Stackelberg Approach

Cloud-Edge-Terminal Collaborative AIGC for Autonomous Driving

Resource Allocation and Data Offloading Strategy for Edge-Computing-Assisted Intelligent Telemedicine System

Edge Intelligence Optimization for Large Language Model Inference with Batching and Quantization