News & Events

Chenguang Wang plans to work with Google DeepMind researchers to improve the grounding ability of large language models with Google Research Scholar Award. (Image created by OpenAI DALL·E)

Consistency, trustworthiness in large language models goal of new research

Chenguang Wang plans to improve grounding ability with Google Research Scholar award.

08.15.2024

WashU researchers have found humans adjust their own behavior to be more fair when training AI.

Humans change their own behavior when training AI

McKelvey’s Chien-Ju Ho working with Art & Sciences’ Wouter Kool, DCDS PhD student Lauren Treiman to understand how human behavior changes in training AI.

08.07.2024

Visual comparison of matched, mismatched and adapted (proposed) deep learning models in PnP-ADMM for the phase retrieval problem. The mismatched model is trained on pathology images instead of faces (matched). The proposed method applies domain adaptation to the mismatched model to restore a high-quality image, comparable in quality to the results achieved using the matched model, with fewer than 1% of the number of images required for training. (Credit: Kamilov lab)

Deep learning models can be trained with limited data

Ulugbek Kamilov, graduate students, develop method that could reduce errors in computational imaging.

07.10.2024

Clockwise from top left: Goodman, Hall, Hu, Kehne, Kong, Lagunas Vargas, Sorrells, Roccabianca, Manzoor, Liu.

Ten new faculty to join McKelvey Engineering for 2024-25

New faculty bring range of experience from data science to biomechanics.

07.02.2024

Satellite images can be synthesized using GeoSynth based on given text prompts. In this example, the model uses a street map to determine the layout and the text prompt “city after earthquake” to fill in the style details of the created satellite image. (Image: Srikumar Sastry)

Artificial intelligence meets cartography

Graduate students in Nathan Jacobs’ lab presented mapping tools to create satellite images from text prompts at EarthVision 2024.

06.17.2024

Using chess as an experimental domain, Chien-Ju Ho will explore how incomplete or imperfect human-labeled data can be used to train superhuman machine learning models. (Credit: Rodrigo Perez on Unsplash)

Can weak human supervision train superhuman models?

Chien-Ju Ho won a grant from OpenAI to study weak human supervision in training powerful machine learning models.

06.13.2024

Researchers who had been using Fitbit data to help predict surgical outcomes have a new method to more accurately gauge how patients may recover from lower back surgery. (Photo: Shutterstock)

New machine learning method can better predict spine surgery outcomes

Chenyang Lu leads collaboration using Fitbit data from patients.

05.30.2024

A new tool developed by computer scientist Ning Zhang embeds distortions imperceptible to human ears into audio recordings to prevent them from being cloned by cybercriminals. (Image generated by Shawn Ballard using Canva)

DeFake tool protects voice recordings from cybercriminals

Ning Zhang is one of three winners of the Federal Trade Commission’s Voice Cloning Challenge to prevent, monitor and evaluate malicious voice cloning.

04.18.2024

The Ursaworks team finished second at the 2024 Midwest Robomaster competition. From left to right: Adheet Ganesh, Xunhao Li, Thomas Upin, Xinshi Feng, Haoyu (Mark) Quan, Thomas Dong, Bruce Li, Steven Luyapan, Erin Chen and Anny Qiao (Photo credit: Mark Quan)

The Ursaworks Robotics Club triumphs on the battlefield

The team has upped its game since a fifth-place showing in 2023.