Justin Bayer

About

Justin Bayer is a researcher whose work bridges the frontiers of machine learning, robotics, and biomechanics. His key contributions span three interconnected domains: deep generative models, human-robot interaction, and probabilistic inference. In machine learning, Bayer made significant strides in generative modeling, with his work on metrics for evaluating deep generative models (36 citations) and variational inference for anomaly detection in high-dimensional time series (40 citations) providing foundational tools for the field. His research on Stochastic Recurrent Networks advanced the ability to model complex spatiotemporal patterns. In robotics and biomechanics, Bayer tackled the challenging problem of human-machine interfaces. His highly cited work on continuous robot control using surface electromyography from atrophic muscles (27 citations) demonstrated practical applications for assistive technologies. He further contributed to understanding hand biomechanics, showing that grip stiffness can be decoupled from force and predicted from electromyography (22 citations)—insights directly applicable to variable-stiffness robotic systems. Bayer also innovated in computer vision, developing methods to estimate fingertip forces from hand images using convolutional neural networks and Gaussian processes (14 citations). His research on empowerment (2 citations) explored fundamental questions about an agent's influence on its environment. Through this diverse body of work, Bayer has made lasting contributions to both the theoretical foundations and practical applications of intelligent systems.

Research Focus

Key Achievements

Top Papers

1. 1
   Variational Inference for On-line Anomaly Detection in High-Dimensional Time Series

   40 citations · 2016
2. 2
   Metrics for Deep Generative Models

   36 citations · 2017
3. 3
   Continuous robot control using surface electromyography of atrophic muscles

   27 citations · 2013
4. 4
   Key Insights into Hand Biomechanics: Human Grip Stiffness Can Be Decoupled from Force by Cocontraction and Predicted from Electromyography

   22 citations · 2017
5. 5
   Estimating finger grip force from an image of the hand using Convolutional Neural Networks and Gaussian processes

   14 citations · 2014
6. 6
   Training Neural Networks with Implicit Variance

   5 citations · 2013
7. 7
   Identification of Human Limb Stiffness in 5 DoF and Estimation via EMG

   3 citations · 2013
8. 8
   Efficient Empowerment

   2 citations · 2015

Key Collaborators

Contact & Links