DeepVerse 6G Machine Learning Challenge
The DeepVerse 6G Machine Learning Challenge is a student competition organized by the Student and Outreach Subcommittee (SOSC) of the IEEE Information Theory Society, in collaboration with the Wireless Intelligence Lab at Arizona State University (ASU) and the Information Theory Labs at National Yang Ming Chiao Tung University. The competition is focused on developing innovative machine learning solutions for various applications using the DeepSense 6G dataset, which comprises coexisting multi-modal sensing and communication data, such as mmWave wireless communication, camera, GPS data, and Radar, collected in realistic wireless environments.
The competition consists of three tasks of increasing difficulty that require participants to use different machine and deep learning modalities and techniques. The scenario considered is as follows: a remote radio head (RRH) is tasked with assisting a base station (BS) in communicating toward a user equipment (UE). The RRH obtains the channel state information (CSI) of the channel between the RRH and UE and wishes to communicate this estimate to the BS through a noiseless but rate-limited channel. The BS reconstructs CSI under a given MSE distortion, while also complementing this estimate through a datastream consisting of radar, GPS, and image data. This CSI estimate is then used in downstream tasks that are not considered further.
Registration Link (Deadline has passed)