Peking University Agent Learning and Intelligence Lab


The lab focuses on reinforcement learning and building generalist agents. Many thanks to our sponsors including NSFC, Huawei, Hikvision, Tencent, inspir.ai, and Alibaba.


Students

Alumni


Current Projects

Generalist Agents

Recently developed foundation models, such as large language models and multi-modal models, open great opportunities to build generally capable agents, combined with reinforcement learning. This project focuses on learning skills and foundation models and connecting them to build generalist agents. In the following, we introduce some of our studies. For more details, please refer to the papers. Plan4MC We study building a multi-task agent in Minecraft. Without human demonstrations, solving long-horizon tasks in this open-ended environment with reinforcement learning (RL) is extremely sample inefficient. [Read More…]

RL/Multi-Agent RL

Multi-Agent Reinforcement Learning (MARL) has recently attracted much attention from the communities of machine learning, artificial intelligence, and multi-agent systems. As an interdisciplinary research field, there are so many unsolved problems, from cooperation to competition, from agent communication to agent modeling, from centralized learning to decentralized learning. MARL has been the main research focus of our lab. We are investigating the field from many perspectives. In the following, we introduce some of our studies. [Read More…]

RL/MARL Applications

Reinforcement learning (RL) has the potential be applied to many real-world applications. In our research, we also investigate the applications of RL and Multi-agent RL. Currently, we have been investigating two applications: one is traffic signal control; another is EDA. Traffic signals coordinating traffic movements are the key for transportation efficiency. However, conventional traffic signal control that heavily relies on pre-defined rules and assumptions on traffic conditions is far from intelligence. [Read More…]


Past Projects

Distributed Video Processing Using Deep Learning on Networked Devices

The vast adoption of mobile devices with cameras has greatly assisted in the proliferation of the creation and distribution of videos. Videos, which are a rich source of information, can be exploited for on-demand information retrieval. Deep learning using Convolutional Neural Networks (CNNs) is state of the art computer vision techniques that can be used for information retrieval. However, due to the high computation of video processing using CNNs, it is not feasible or costs too much to process all videos at a centralized entity, considering a large set of videos which is common in this big data epoch. [Read More…]

Building Smartphone Networks

Smartphones have great networking capabilities. They can access the Internet through cellular networks or wireless access points and communicate with nearby devices using WiFi Direct or Bluetooth. However, these network functions may not work in some circumstances where cellular towers and network infrastructure are destroyed, e.g. in disaster recovery. Nevertheless, communications in such scenarios are very important, and hence, in this research, we aim to build smartphone networks to provide communications without relying on cellular networks, wireless access points, or network infrastructure. [Read More…]

Health Sensing Using Mobile Devices

Mobile devices, such as smartphones, have become commonplace in health care settings, leading to the development of both platforms and applications for health care, e.g., HealthKit on iOS, where apps can collect users’ health and activity data and the data will be used for medical research to bring more powerful health solutions. However, no data is collected for the research of infectious diseases. Moreover, currently, most health data are collected by manual input or external devices. [Read More…]

Exploring Social Structure for Network Designs

The proliferation of mobile devices, such as smartphones and tablets, and the popularity of online social networks that link humans, mobile devices and Internet together, increasingly emphasize the role of human behaviors on network designs. Due to the involvement of human behaviors, social structure provides crucial information of network structure and node organization, and thus can be exploited for network designs, e.g., in online social networks and mobile social networks. [Read More…]